Sopk Fezolinetant
Sopk Fezolinetant
Graeme L. Fraser,1 Barbara Obermayer-Pietsch,2 Joop Laven,3 Georg Griesinger,4 Axelle Pintiaux,5
Dirk Timmerman,6 Bart C.J.M. Fauser,7 Christopher Lademacher,8 Jean Combalbert,1 Hamid R.
Hoveyda,1 and Steven Ramael1
t
Hospital, Lubeck, Germany; 5CHU Liège Citadelle Hospital, Liege, Belgium; 6University Hospital KU
ip
Leuven, Leuven, Belgium; 7University Medical Center, Utrecht, Netherlands; 8Astellas Pharma, Inc.,
Chicago, IL, USA
cr
us
ORCiD numbers:
Endocrine Society.
This is an Open Access article distributed under the terms of the Creative Commons
Attribution-NonCommercial-NoDerivs licence
reproduction and distribution of the work, in any medium, provided the original work
is not altered or transformed in any way, and that the work is properly cited. For
EPICS Therapeutics
t
E-mail: [email protected]
ip
cr
Financial Support: This study was sponsored by OGEDA S.A. (Gosselies, Belgium), a wholly owned
subsidiary of Astellas Pharma, Inc. Medical writing and editorial support were provided by Lauren A.
us
Cerruto and Diane M. Sloan, PharmD, of Peloton Advantage, LLC (Parsippany, NJ), an OPEN Health
company, and funded by Astellas Pharma, Inc.
an
Disclosure Summary: G.L.F. has served as a consultant to Astellas Pharma, was
M
Astellas Pharma. B.O-P. has nothing to disclose. J.L. received grants from Astellas
e
pt
Pharma during the conduct of the study and has received grants and personal fees
from Ansh Labs and Ferring, personal fees from Danone and Titus Healthcare, and
ce
grants from Merck Serono and ZonMw. G.G. received compensation for the conduct
Ac
of the study and has received honoraria and/or nonfinancial support from Abbott,
Ferring, Finox, Gedeon Richter, Glycotope GmbH, Guerbet, Merck Serono, MSD,
University Hospital Liège) from Astellas Pharma during the conduct of the study and
is a consultant for Bayer, Ceres, and Gedeon Richter. D.T. received fees (paid to
University Hospitals Leuven) from Astellas Pharma during the conduct of the study.
2
B.C.J.M.F. has received fees or grant support from the following organizations (in
t
ip
PregLem/Gedeon Richter, Reproductive Biomedicine Online (RBMO), and World
cr
Health Organisation (WHO). C.L. is an employee of Astellas Pharma. J.C. is an
employee of EPICS Therapeutics. H.R.H. and S.R. are former employees of OGEDA
SA.
us
an
Email addresses
M
B. Obermayer-Pietsch: [email protected]
d
J. Laven: [email protected]
e
G. Griesinger: [email protected]
pt
A. Pintiaux: [email protected]
ce
D. Timmerman: [email protected]
C. Lademacher: [email protected]
J. Combalbert: [email protected]
S. Ramael: [email protected]
3
ABSTRACT
Context: Polycystic ovary syndrome (PCOS), a highly prevalent endocrine disorder characterized by
hyperandrogenism, is the leading cause of anovulatory infertility.
Objective: This proof-of-concept study evaluated clinical efficacy and safety of the neurokinin 3
(NK3) receptor antagonist fezolinetant in PCOS.
Design: This was a phase 2a, randomized, double-blind, placebo-controlled, multicenter study
t
Patients: Women with PCOS participated in the study.
ip
Intervention: Interventions included fezolinetant 60 or 180 mg/d or placebo for 12 weeks.
cr
Main Outcome Measure: The primary efficacy endpoint was change in total testosterone.
Gonadotropins, ovarian hormones, and safety/tolerability were also assessed.
us
Results: Seventy-three women were randomized, and 64 participants completed the study. Adjusted
mean (SE) changes in total testosterone from baseline to week 12 for fezolinetant 180 and 60 mg/d
an
were −0.80 (0.13) and −0.39 (0.12) nmol/L versus −0.05 (0.10) nmol/L with placebo (P<0.0001 and
P<0.05, respectively). Adjusted mean (SE) changes from baseline in luteinizing hormone (LH) for
fezolinetant 180 and 60 mg/d were −10.17 (1.28) and −8.21 (1.18) versus −3.16 (1.04) IU/L with
M
placebo (P<0.0001 and P=0.0022); corresponding changes in follicle-stimulating hormone (FSH) were
−1.46 (0.32) and −0.92 (0.30) versus −0.57 (0.26) IU/L (P=0.0336 and P=0.3770), underpinning a
dose-dependent decrease in the LH-to-FSH ratio versus placebo (P<0.001). Circulating levels of
d
progesterone and estradiol did not change significantly versus placebo (P>0.1). Fezolinetant was well
tolerated.
e
Conclusions: Fezolinetant had a sustained effect to suppress hyperandrogenism and reduce the LH-
pt
to-FSH ra
ce
4
INTRODUCTION
Polycystic ovary syndrome (PCOS) is the leading cause of anovulatory infertility and the most
common endocrine disorder in reproductive-aged women (1), with an estimated global prevalence
of approximately 10% (2) and related annual medical costs exceeding $4 billion in the United States
(3, 4). PCOS diagnostic criteria include clinical or biochemical hyperandrogenism, chronic oligo-
ovulation or anovulation, and polycystic ovaries (5, 6). Diagnosis depends on identifying 2 of these 3
phenotypic features in the absence of other etiologies (Rotterdam criteria (6)). Metabolic features of
t
ip
with complex symptomology, and the etiology of the disease remains unclear.
cr
Altered signaling in the neuroendocrine circuits that regulate fertility is considered to be a
preponderant feature of PCOS (10). The hypothalamic network of kisspeptin, neurokinin B, and
us
dynorphin A (KNDy) neurons has been identified as the gonadotropin-releasing hormone (GnRH)
pulse generator that governs the pattern of luteinizing hormone (LH) and follicle-stimulating
hormone (FSH) secretion over the phases of the ovarian cycle (11-13). Patients with PCOS often
an
express high-frequency pulses of LH, increased serum LH, and a high LH-to-FSH ratio (14-16). At the
level of the ovary, high LH increases androgen synthesis, whereas (relatively) low FSH may
contribute to follicular arrest, anovulation, and accumulation of cysts (10). Moreover, the androgen
M
excess contributes to the impaired negative feedback of ovarian hormones on the LH pulse
frequency and thereby fuels an arrhythmic reproductive cycle (17-20).
e d
The current standard of care for PCOS is treatment with hormone contraceptives for managing
pt
menstrual irregularities and certain symptoms of hyperandrogenism (i.e., hirsutism, acne, and
alopecia) (5, 21, 22). The estrogen component of hormone contraceptives increases sex hormone–
binding globulin (SHBG) and reduces LH and FSH, resulting in a decrease in androgen production and
ce
therapy in women with PCOS who present with abnormal glucose tolerance or type 2 diabetes (5,
22). Spironolactone, a potassium-sparing diuretic with antiandrogen properties, is sometimes used
in combination with hormone contraceptives to help alleviate the manifestations of
hyperandrogenism (22, 24). In summary, all current treatments are aimed at ameliorating symptoms
and correcting the biochemical imbalance of PCOS but do not address the central hormonal
dysregulation.
Neurokinin 3 (NK3) receptor signaling has been shown to play a key role in positive and negative
feedback loops regulating the hypothalamic-pituitary-gonadal (HPG) axis (25, 26). In premenopausal
5
women, NK3 receptor antagonism at the level of the KNDy neuron is understood to decrease the
GnRH pulse frequency based on the downstream observations of reduced basal LH secretion, lower
LH-to-FSH ratio, suppressed follicle development, and the modulation of the temporal dynamics of
ovarian sex hormone production over the menstrual cycle (25-27). The pharmacology of NK3
receptor antagonists in the regulation of the HPG axis inspired us to investigate whether such
compounds could correct the elevated GnRH pulse frequency attributed to the neuroendocrine
impairments associated with PCOS and thereby improve clinical outcomes. In the interim, the NK3
receptor antagonist MLE4901 (formerly AZD4901) was investigated in exploratory phase 2 studies
t
ip
cr
Fezolinetant (ESN364) is a novel oral small molecule that potently and selectively blocks the NK3
receptor (31). Preclinical data demonstrated that administration of fezolinetant decreased LH pulse
us
frequency and lowered plasma LH without affecting FSH (32). Fezolinetant treatment for 21 days
produced dose-dependent decreases in LH with no significant effect on FSH, leading to decreases in
the LH-to-FSH ratio in healthy female volunteers with regular ovulatory menstrual cycles (26).
an
Fezolinetant is now in phase 3 development for treatment of vasomotor symptoms in
postmenopausal women, following promising efficacy and safety data for this indication in two
phase 2 trials (33, 34). The current study was conducted to evaluate the effects of fezolinetant on
M
This phase 2, proof-of-concept, randomized, double-blind, multicenter study evaluated the efficacy
ce
and safety of fezolinetant versus placebo administered for 12 weeks in women with PCOS (EudraCT
Number: 2014-004409-34). The study was conducted entirely at academic or clinical (hospital) sites
from May 2015 through May 2017 in 5 European countries (Austria, Belgium, Georgia, Germany, and
Ac
Netherlands).
The study included a screening period (−28 to −7 days before first dose), during which baseline data
were collected. Eligible women then entered a 12-week, double-blind, placebo-controlled treatment
period and were randomized 1:1:1 via computer-generated randomization schedule to receive
fezolinetant 60 mg, fezolinetant 180 mg, or matching placebo. All study drugs were administered
orally once daily after a light breakfast for up to 12 weeks. Patients visited the clinical center every 3
weeks for assessments and attended a follow-up visit 6 weeks after completing treatment. All in-
study visits were to be planned within 2 to 8 hours after study drug intake, except for visit 2
(randomization visit; week 1, day 1) and visit 5 (week 9, day 63). These two visits were done with
6
patients in the fasted state in the morning to evaluate baseline parameters and, in the case of visit 5,
to measure trough pharmacokinetics (PK) levels and hormonal effects at trough PK levels.
Ethical considerations
The study protocol was reviewed and approved by an independent ethics committee and/or
institutional review board at each study site, and the study was conducted in accordance with the
t
ip
Study population
cr
Patients were women aged 18 to 45 years with a diagnosis of PCOS according to the Rotterdam
criteria (6), with the modification of mandatory biochemical hyperandrogenism (total testosterone:
us
>1.7 nmol/L). At least 1 of the following 2 other Rotterdam criteria were also required for diagnosis
of PCOS: oligomenorrhea (≤6 menses per year) or oligo-ovulation and/or polycystic ovaries on
ultrasound scan (at least 1 ovary with ≥12 antral follicles or ovarian volume ≥10 cm3). Additional
an
inclusion criteria were normal thyroid function; normal levels of FSH, estradiol (E2), prolactin, and
17-hydroxyprogesterone; and good physical and mental health based on medical history and
M
examination. Patients were also required to have negative cervical cytology within 36 months of
screening, a negative urine test for drugs of abuse, and a negative pregnancy test and were required
to use highly effective nonhormonal contraception through 42 days posttreatment if sexually active.
e d
Exclusion criteria included evidence of diabetes based on World Health Organization criteria (35, 36);
pt
aminotransferase (ALT) or aspartate aminotransferase levels >1.3 times the upper limit of normal
(ULN), total bilirubin >1.3 times the ULN, or creatinine >1.25 times the ULN; hemoglobin <10 g/dL;
Ac
positive hepatitis panel or HIV antibody test at screening; psychological disorder within 1 year before
screening; symptomatic acute or chronic illness within 3 months of initial study drug administration;
and significant blood loss or transfusion within 12 weeks of study drug administration.
Patients were excluded if they had received any of the following within 3 months before screening:
antiandrogens, GnRH agonist/antagonists, selective estrogen receptor modulators, selective
progesterone (P4) receptor modulators, dienogest, danazol, aromatase inhibitors, glucocorticoids,
mineralocorticoids, androgens, or depot contraceptives. Any hormonal contraceptives were required
to be stopped 1 month prior to screening, and any insulin sensitizers discontinued at screening. Any
patient deemed by the investigator to be inappropriate for the study based on electrocardiographic
7
abnormalities or acute or chronic medical condition that could either interfere with drug PK or
interpretation of the study outcomes was excluded.
The primary efficacy endpoint was mean change in total testosterone from baseline to week 12 (end
of treatment). Secondary efficacy endpoints included changes in levels of other gonadotropins and
t
Syndrome Questionnaire (PCOSQ) score (37, 38); and changes in transvaginal ultrasound parameters
ip
(endometrial thickness, ovarian volume, number of follicles [cysts], and surface of the dominant
follicle) from baseline to weeks 6 and 12.
cr
us
Analysis of total testosterone was performed on frozen (−20°C) plasma samples by SGS Life Sciences,
Wavre, Belgium, using a validated liquid chromatography with tandem mass spectrometry method
(39). The lower limit of quantitation was 25.0 pg/mL, percent coefficient of variation ranged from
an
1.41% to 2.44%, and percent relative error ranged from −3.77% to 2.08%. Other gonadotropin and
ovarian hormone analyses were performed using validated analytical methods on frozen (<−70°C)
plasma samples by BARC, Ghent, Belgium (LH, FSH, and P4) and CERBA, Paris, France (E2).
M
To allow assessment of changes in menstrual cycle, patients recorded start and end dates of any
d
vaginal bleeding, as well as severity (none, spotting, light, normal, or heavy), in an electronic diary.
They also completed the PCOSQ in an electronic diary every 3 weeks through week 12 and then at
e
questions in 5 domains: emotional, body hair, infertility, weight, and menstrual problems scored on
a scale of 1 to 7 (37, 38). Two-dimensional transvaginal ultrasound was performed at screening,
ce
baseline, and weeks 6 and 12 using standardized instrument settings for beam focus, overall time-
gain, and near-field and far-field gain. Central reading of the ultrasonography for ovarian volume,
endometrial thickness, number of follicles, and dominant follicle development was performed by an
Ac
8
analyzed using a validated liquid chromatography with tandem mass spectrometry method. The
assay had a quantification limit of 5.00 ng/mL, a percent coefficient of variation of 5.51% to 7.65%,
and relative errors of −4.60% to 0.75%.
Exploratory pharmacodynamic endpoints, evaluated by the BARC laboratory using frozen (<−70°C)
plasma samples at baseline and weeks 6, 12, and 18, included changes in levels of leptin,
androstenedione, aldosterone, dehydroepiandrosterone sulfate, SHBG, anti-Müllerian hormone
t
Safety endpoints included adverse event (AE) frequency and severity, hematology and biochemistry
ip
assessments, changes in levels of bone density markers (bone alkaline phosphatase and beta-
carboxy-terminal peptide of type I collagen), and change in Columbia-Suicide Severity Rating Scale
cr
(C-SSRS) score from baseline to weeks 12 (end of treatment) and 18 (follow-up) (40, 41). Safety
assessments also included physical examination, hematology and biochemistry testing, vital signs,
us
and electrocardiographic findings.
an
Statistical analysis
Planned total enrollment was 72 patients, with 24 randomized to each treatment group; no formal
M
power calculations were made. The safety population included all randomized patients who received
at least one dose of study medication. The intent-to-treat population included patients from the
safety population who had at least one postbaseline efficacy assessment.
e d
Statistical calculations were performed by SGS Life Sciences using SAS version 9.2 or higher (SAS
pt
Institute, Inc., Cary, NC, USA) and/or WinNonlin version 5.2 or higher software (Pharsight Corp.,
Mountain View, CA, USA). All endpoints were summarized descriptively. A post hoc analysis was
ce
performed for sex steroid hormone parameters using an analysis of covariance model with
treatment group as a fixed factor and baseline value as a covariate. A possible relationship between
drug-plasma concentrations and clinical response, as measured by changes in LH, FSH, LH-to-FSH
Ac
ratio, P4, E2, and total testosterone concentrations, was graphically explored. For safety endpoints,
frequencies of AEs were tabulated and analyzed in a descriptive manner, with AEs coded according
to the Medical Dictionary for Regulatory Activities version 18.0.
9
RESULTS
Of 105 patients screened, 73 were randomized and included in the safety and intent-to-treat
populations and 64 completed the study (Figure 1). Treatment groups were well matched for
demographics and baseline clinical characteristics (Table 1).
Both doses of fezolinetant significantly reduced total testosterone relative to placebo at week 12
t
(Table 2), the primary endpoint. A dose-related effect was seen such that the 180 mg dose
ip
significantly (P<0.001) reduced total testosterone relative to placebo at all timepoints during
treatment, whereas the 60 mg dose significantly (P<0.05) reduced total testosterone at only weeks 3
cr
and 12. Specifically, change (95% confidence intervals [CIs]) from baseline in total testosterone at
weeks 3, 6, and 12 were −19% (−28.7, −9.0), −14% (−25.6, −1.8), and −17% (−28.7, −4.6),
us
respectively, with fezolinetant 60 mg; −32% (−42.1, −21.3), −31% (−43.9, −17.7), and −33% (−45.91,
−20.4), respectively, with fezolinetant 180 mg; and −3% (−11.1, 5.7), −1% (−11.4, 9.4), and 1% (−8.8,
11.7), respectively, with placebo (Figure 2A). The reduction (95% CI) from baseline in total
an
testosterone at trough PK concentrations (week 9) was 8% (−18.0, 2.3) with 60 mg and 24% (−35.5,
−11.7) with 180 mg, the latter was also significant versus placebo (P=0.005), indicating that
fezolinetant 180 mg suppresses androgen throughout the day.
M
Changes in LH and FSH at week 12 are shown in Table 2 and Figure 2B and 2C respectively. Both
doses of fezolinetant significantly reduced concentrations of LH to a greater extent than FSH,
pt
thereby significantly (P<0.001) decreasing the LH-to-FSH ratio relative to placebo at week 12. A dose-
dependent effect was also seen in reductions in the LH-to-FSH ratio, which were better sustained
ce
with fezolinetant 180 mg, especially at trough PK concentrations (week 9) (Figure 2D).
Ac
For both doses of fezolinetant, changes in E2 and P4 concentrations at week 12 were not
significantly different from placebo (Table 2). Fezolinetant 180 mg reduced E2 from baseline, but
these changes were not significant at weeks 3, 6, or 12 compared with changes seen with placebo
(Figure 3A). As shown in Table 2, P4 sampling on the same schedule indicated a tendency for
fezolinetant 180 mg to reduce P4 from baseline to week 12, but this change was not statistically
significant. Ten patients (placebo: n=7; fezolinetant 60 mg: n=3; fezolinetant 180 mg: n=0) had P4
concentrations >6.0 ng/mL at any time during active treatment, indicative of ovulation (42);
however, the small number of patients and sporadic timing of these elevated P4 readings precludes
any clear relationship to treatment.
10
Clinical outcomes
Fezolinetant was not associated with clinically meaningful changes in PCOSQ scores (Table 3). On
transvaginal ultrasound examinations, endometrial thickness for both fezolinetant groups was
similar to or lower than for the placebo group throughout the treatment duration (Figure 3B).
Treatment with fezolinetant did not regularize the menstrual cycles (Figure 3C).
A decrease in AMH over the duration of the study was observed in the 180 mg group, although this
finding was not statistically significant (Figure 4A). Change from baseline in ovarian volume based on
t
transvaginal ultrasound was not significantly different for fezolinetant versus placebo, although total
ip
ovarian volume trended downward at week 12 in the fezolinetant 180 mg group (Figure 4B). There
were no significant changes observed in the number of follicles nor surface size of the dominant
cr
follicle (Table 4).
180 mg group, which correlated with the changes in total testosterone levels.
Ac
Safety
Overall, treatment with fezolinetant for 12 weeks was safe and well tolerated. Treatment-emergent
AEs (TEAEs) occurring in at least 3 patients are listed in Table 6. TEAEs that occurred in at least 3
patients exposed to fezolinetant were headache, paresthesia, rash, nausea, and nasopharyngitis. The
most frequently reported TEAEs considered at least possibly related to treatment by the investigator
were headache (placebo: 18.5%, fezolinetant 60 mg: 13.0%, and fezolinetant 180 mg: 39.1%) and
paresthesia (0%, 0%, and 17.4%, respectively).
11
Three serious TEAEs were reported in 3 fezolinetant-treated patients. One patient in the fezolinetant
60 mg group experienced superficial thrombophlebitis that was determined to be possibly related to
treatment; the drug was temporarily stopped and then restarted without incident. The other two
serious TEAEs, both in the 180 mg group, were an ankle fracture associated with a horse riding
accident that required temporary study drug discontinuation during surgery (n=1) and severe
sciatica (n=1), neither of which was deemed related to treatment. All other TEAEs were mild to
One patient in the fezolinetant 180 mg group discontinued study drug because of AEs of depressed
t
ip
mood, headache, decreased libido, and mood swings. These AEs were determined by the study
sponsor to be potentially related to treatment, and the study drug was permanently discontinued.
cr
us
No clinically relevant changes in clinical laboratory parameters, vital signs, electrocardiographic
values, or bone density markers were observed. The most frequently reported treatment-emergent
laboratory abnormalities (observed in ≥4 patients in any treatment group) were elevated ALT,
an
calcium, creatinine, urate, and hemoglobin and lower than normal levels of leukocytes and
neutrophils. There was an apparent dose-related increase in creatinine seen in 3.7% of patients in
the placebo group, 4.3% of the fezolinetant 60 mg group, and 18.2% of the fezolinetant 180 mg
M
group.
d
Treatment-emergent ALT increases, based on laboratory monitoring, were equally distributed over
the treatment groups. All reported increases were <3 x ULN except for in 1 patient in the placebo
e
group who had ALT values of 5.5 × ULN at week 12 and 1 patient in the 180 mg group with an ALT of
pt
3.2 x ULN at week 3. All increases were transient and resolved spontaneously. In the 180 mg group, 1
patient had an ALT of 8.6 × ULN at week 3, but her baseline value was 14.3 × ULN. Thus, the on-
treatment ALT elevation was a pre-existing condition, and because her liver test values normalized
ce
during treatment and in the absence of any concomitant signs or symptoms, this event was not
regarded as treatment-emergent.
Ac
C-SSRS scores were negative in all but 1 patient who developed moderate treatment-emergent
depression possibly related to treatment; the patient recovered from this TEAE by day 61 with a
negative C-SSRS score at follow-up.
DISCUSSION
This randomized, double-blind, placebo-controlled study supports the concept that NK3 antagonist
therapy offers potential benefit in the treatment of PCOS. Fezolinetant produced significant
reductions in hyperandrogenemia during 12 weeks of treatment in women with PCOS, reducing total
testosterone with daily doses of 60 and 180 mg by 14% to 19% and 31% to 33% at peak drug
12
concentrations, for each dose respectively. Fezolinetant 180 mg produced a sustained reduction of
total testosterone over the 24-hour dose interval (24% at trough drug concentrations [P=0.005]) and
also consistently lowered the LH-to-FSH ratio to within a normal range such that LH and FSH, as well
as estrogen, were maintained at levels comparable to those in healthy women in the early to mid-
follicular phase (15, 26).
Results are consistent with the hypothesis that antagonism of NK3 receptor signaling affects the
t
production of sex hormones by the ovaries (26, 32). The GnRH pulse generator is regulated by
ip
negative feedback from the ovarian hormones P4 and E2 (44). Across the range of clinical studies
conducted to date, it is interesting to observe similar sensitivity in response to fezolinetant on the
cr
gonadotropins, LH and FSH, under conditions of normal feedback from ovarian hormones (26),
negligible feedback from ovarian hormones (e.g., menopause (33)), and disordered feedback from
us
ovarian hormones (e.g., PCOS (18)), as shown here. These data provide further empirical evidence
for the functional hierarchy of hypothalamic NK3 over kisspeptin signalling pathways in mediating
feedback mechanisms on the HPG axis (25).
an
Exogenous administration of a kisspeptin agonist does not change the dynamics of LH secretion in
M
PCOS patients relative to healthy controls (45), indicating that any neuroendocrine basis of PCOS is
at a hierarchical level above that of kisspeptin signalling. Thus, altered KNDy neuron signalling and
resultant changes in the pattern of endogenous kisspeptin secretion may be relevant to disease. In
d
the current study, fezolinetant significantly decreased FSH in patients with PCOS, as previously
observed only during the mid-cycle gonadotrophin surge in premenopausal healthy female
e
volunteers (26). Thus, perhaps the distinct neuroendocrine dynamics relevant to surge (13) also
pt
pertain to the etiology of PCOS. In primates, the surge generator is reliant upon estrogen positive
feedback at the level of the pituitary and/or involves distinct neuronal circuits in the mediobasal
ce
hypothalamus (44, 46). Although kisspeptin is involved in estrogen-induced surge, the specific role of
KNDy neurons in this phenomenon is unclear (47). Notably, it is only under estrogen positive
feedback that NK3 antagonist treatment improved the regularity and orderliness of kisspeptin-
Ac
The sustained reduction in total testosterone and LH over the entire treated population in the
current 12-week study compares well with the preliminary findings from a trial of MLE4901
(formerly AZD4901), another NK3 antagonist, in which similar findings were apparent in all patients
at day 7 but only in suspected anovulatory patients at day 28 (28). In this previous study, the authors
acknowledge that P4 concentrations in the treatment phase were higher than expected for PCOS
patients in general, and the suspected ovulatory patients (i.e., those with P4 >6 ng/mL (42)) were
13
removed from post hoc analyses (28). In contrast, our findings indicate that NK3 antagonist
treatment tends to lower the incidence of P4 elevation and subsequent post hoc analyses were not
obliged. MLE4901 has since been discontinued, as the risk/benefit profile no longer supported
continued development (30). The duration of this initial, exploratory study with MLE4901 was
acknowledged to be insufficient to assess clinical outcomes, and such findings were not reported
(28).
t
improvement in menstrual cycle regularity, follicle counts, or PCOSQ scores. The 12-week duration
ip
of treatment in this trial may be inadequate to affect these parameters as positive clinical outcomes
in PCOS clinical trials are typically detected after 6 to 9 months of treatment (49, 50). However, the
cr
small but consistent, dose-related decreases in serum AMH levels, together with the associated
decreases in ovarian volume (51-53), may be an early sign of improved clinical outcomes with
us
prolonged treatment. The time course required to lower AMH levels could relate to the necessity to
replace the whole follicle cohort against a background of reduced androgen (50).
an
Diverse approaches ranging from retrospective clinical studies to mathematical modeling converge
to conclude that androgenic effects at both the pituitary and the ovaries contribute to the etiology
M
of PCOS (54, 55). High total testosterone correlates with a larger number of small antral follicles (56)
leading to increased AMH production (57) and follicular development arrest (58). AMH also inhibits
cytochrome P450 aromatase (CYP19, the enzyme that converts androgens to E2) in the granulosa
d
Furthermore, both total testosterone and AMH modulate hypothalamic-pituitary circuits to elevate
the LH-to-FSH ratio. A positive correlation between LH and AMH serum levels is evident in women
ce
with PCOS (53), a finding consistent with recent mechanistic studies in rodent models demonstrating
that AMH has direct, positive feedback on GnRH neuronal activation to increase LH pulsatility (60,
61). Also, genome-wide association analyses conclude that testosterone has an etiologic role in PCOS
Ac
(62) in accordance with mechanistic studies demonstrating that elevated androgens affect the
plasticity of key neuroendocrine circuits (63, 64). The latter finding is consistent with the
pharmacologic demonstration that long-term androgen receptor blockade restores the negative
feedback of estrogen and P4 on LH pulse frequency in women with PCOS (19). Thus, the trends
toward lower levels of both total testosterone and AMH shown here in response to fezolinetant
suggest that longer-term studies would be of interest to confirm whether improved clinical
outcomes are achievable.
This study has focused on the neuroendocrine axis as the target site for interpreting NK3 antagonist
effects in PCOS. However, the NKB-NK3 receptor signaling pathway is also present in human
14
granulosa and cumulus cells (65), where a significant decrease in NK3 receptor expression (e.g.,
TACR3 mRNA) (66-68) may contribute to the decreased aromatase levels in women with PCOS (69).
Any direct actions of fezolinetant at the level of the ovary were not evaluated in this trial.
t
volunteers and postmenopausal women (26, 33). The stable exposure levels over the duration of the
ip
study indicate that there is no drug accumulation or modulation of PK processes consequent to
repeated dosing.
cr
us
Fezolinetant was generally well tolerated in women with PCOS. The only serious, potentially
treatment-related AE was superficial thrombophlebitis, which occurred in a single patient in the
fezolinetant 60 mg group. One patient in the 180 mg dose group discontinued treatment because of
an
mood-related effects considered to be potentially treatment related, and one patient had a
transiently positive C-SSRS score. No patients discontinued treatment because of elevated liver
enzymes, and the incidence of treatment-emergent increases in liver enzymes were similarly
M
Women with PCOS may have an increased risk of endometrial cancer (70). Endometrial health was
e
assessed and, as previously demonstrated in premenopausal and menopausal women (26, 34),
fezolinetant treatment had no estrogen-like stimulatory or proliferative effects on the endometrium.
pt
ce
Limitations of this proof-of-concept study include the exploratory nature of the statistical analyses,
the focus on biochemical biomarkers (e.g., total testosterone, LH-to-FSH ratio) over clinical
outcomes as influenced by the short duration of the study, the sufficiency of the 1-month stop
Ac
interval of any oral contraceptives prior to screening, and the small sample size. Although a
hyperandrogenemic population was selected, other factors (e.g., incidence of oligomenorrhea,
metabolic markers, BMI, SHBG) potentially relevant to PCOS patient clustering were uncontrolled in
this study such that heterogeneity in the study population may be a confounding factor in the
interpretation of results (8, 9). At the time that this study was launched, there was no precedence
for this type of therapy in a PCOS trial, and therefore no formal power calculation was performed to
define the sample size. However, an estimate of sample size was made based on assumptions
regarding a projected effect size of 30% to lower both total testosterone and LH on the basis of the
response to fezolinetant measured in healthy volunteers (26).
15
In conclusion, this is the first study to demonstrate that an NK3 receptor antagonist has a sustained
effect in women with PCOS to normalize the LH-to-FSH ratio and reduce the hyperandrogenic state.
However, these changes in hormones did not translate into improved clinical outcomes in this 12-
week study. These data suggest that therapy principally targeting the hypothalamic KNDy-HPG axis
elicits positive changes in biochemical biomarkers but that the expected, consequent changes in the
plasticity of relevant, neuroendocrine circuits and ovarian physiology correct slowly; therefore,
t
ip
cr
us
an
M
e d
pt
ce
Ac
16
ACKNOWLEDGMENTS
The authors thank all members of the OGEDA Drug Discovery Team Astellas Pharma, Inc., acquired
100% of the equity of OGEDA SA on May 17, 2017. Development of this manuscript, including
editorial support provided by Lauren Cerruto and Traci Stuve, MA, of Echelon Brand
Communications (Parsippany, NJ, USA), an OPEN Health company, was sponsored by Astellas
Pharma Global Development. The authors would like to thank all members of the OGEDA Drug
Discovery Team; Joyce Kingsbury, Principal Biostatistician and PHASTAR (Derry, NH, USA), and
t
ip
Financial Support
cr
Clinical Trial Information
The datasets generated during and/or analyzed during the current study are not publicly
available but are available from the corresponding author on reasonable request.
e d
Employees of OGEDA SA (Gosselies, Belgium) were involved in the design and conduct of the study;
collection, management, analysis, and interpretation of the data; preparation, review, or approval of
ce
the manuscript; and decision to submit the manuscript for publication. Employees of Astellas
Pharma, Inc., were involved in the analysis, and interpretation of the data; preparation, review, or
approval of the manuscript; and decision to submit the manuscript for publication.
Ac
17
REFERENCES
t
ip
burden of the polycystic ovary syndrome during the reproductive life span. J
cr
Clin Endocrinol Metab. 2005;90(8):4650-4658.
4. Jason J. Polycystic ovary syndrome in the United States: clinical visit rates,
us
characteristics, and associated health care costs. Arch Intern Med.
an
2011;171(13):1209-1211.
5. Legro RS, Arslanian SA, Ehrmann DA, et al. Diagnosis and treatment of
M
2016;106(1):6-15.
8. Huang CC, Tien YJ, Chen MJ, Chen CH, Ho HN, Yang YS. Symptom patterns
Reprod. 2015;30(4):937-946.
18
9. Dapas M, Lin FTJ, Nadkarni GN, et al. Distinct subtypes of polycystic ovary
t
ip
12. Lippincott MF, Leon S, Chan YM, et al. Hypothalamic reproductive endocrine
cr
pulse generator activity independent of neurokinin B and dynorphin signaling. J
us
13. Herbison AE. A simple model of estrous cycle negative and positive feedback
an
regulation of GnRH secretion. Front Neuroendocrinol. 2020:100837.
14. Blank SK, McCartney CR, Marshall JC. The origins and sequelae of abnormal
M
2006;12(4):351-361.
e d
15. Waldstreicher J, Santoro NF, Hall JE, Filicori M, Crowley WF, Jr. Hyperfunction
pt
Metab. 1988;66(1):165-172.
Ac
16. Taylor AE, McCourt B, Martin KA, et al. Determinants of abnormal gonadotropin
17. Daniels TL, Berga SL. Resistance of gonadotropin releasing hormone drive to
19
18. Pastor CL, Griffin-Korf ML, Aloi JA, Evans WS, Marshall JC. Polycystic ovary
19. Eagleson CA, Gingrich MB, Pastor CL, et al. Polycystic ovarian syndrome:
t
ip
Endocrinol Metab. 2000;85(11):4047-4052.
cr
20. Chang RJ. The reproductive phenotype in polycystic ovary syndrome. Nat Clin
us
21. Barthelmess EK, Naz RK. Polycystic ovary syndrome: current status and future
an
perspective. Front Biosci (Elite ed). 2014;6:104-119.
https://www.monash.edu/__data/assets/pdf_file/0004/1412644/PCOS_Evidenc
e-Based-Guidelines_20181009.pdf.
ce
23. Briden L, Shirin S, Prior JC. The central role of ovulatory disturbances in the
Ac
treatment with cyclic progesterone. Drug Discov Today Dis Models. 2020;32:71-
82.
24. Sirmans SM, Pate KA. Epidemiology, diagnosis, and management of polycystic
20
25. Skorupskaite K, George JT, Veldhuis JD, Millar RP, Anderson RA. Interactions
26. Fraser GL, Ramael S, Hoveyda HR, Gheyle L, Combalbert J. The NK3 receptor
27. Skorupskaite K, George JT, Veldhuis JD, Anderson RA. Neurokinin B regulates
t
ip
gonadotropin secretion, ovarian follicle growth, and the timing of ovulation in
cr
healthy women. J Clin Endocrinol Metab. 2018;103(1):95-104.
us
women with polycystic ovary syndrome: a randomized, placebo-controlled trial.
an
J Clin Endocrinol Metab. 2016;101(11):4313-4321.
29. Skorupskaite K, George JT, Veldhuis JD, Millar RP, Anderson RA. Kisspeptin
M
30. Modi M, Dhillo WS. Neurokinin B and neurokinin-3 receptor signaling: promising
pt
Med. 2019;37(3):125-130.
ce
31. Hoveyda HR, Fraser GL, Dutheuil G, et al. Optimization of novel antagonists to
Ac
the neurokinin-3 receptor for the treatment of sex-hormone disorders (part II).
32. Fraser GL, Hoveyda HR, Clarke IJ, et al. The NK3 receptor antagonist ESN364
21
33. Depypere H, Timmerman D, Donders G, et al. Treatment of menopausal
34. Fraser GL, Lederman S, Waldbaum A, et al. A phase 2b, randomized, placebo-
Menopause. 2020;27(4):382-392.
t
ip
35. World Health Organization. Definition and diagnosis of diabetes mellitus and
cr
intermediate hyperglycemia: report of a WHO/IDF consultation. Geneva,
us
http://www.who.int/diabetes/publications/Definition%20and%20diagnosis%20of
an
%20diabetes_new.pdf.
http://www.who.int/diabetes/publications/report-hba1c_2011.pdf.
pt
life questionnaire (PCOSQ) for women with polycystic ovary syndrome (PCOS).
ce
38. Jones GL, Benes K, Clark TL, et al. The Polycystic Ovary Syndrome Health-
2004;19(2):371-377.
39. SGS Life Science Services. Validation of a LC/MS-MS method for the
22
report no. B1121044]. Wavre, Belgium: SGS Life Science, division of SGS
40. Posner K, Brown GK, Stanley B, et al. The Columbia-Suicide Severity Rating
Scale: initial validity and internal consistency findings from three multisite
t
ip
http://cssrs.columbia.edu/the-columbia-scale-c-ssrs/about-the-scale/.
cr
42. Welt CK. Evaluation of the menstrual cycle and timing of ovulation. UpToDate;
menstrual-cycle-and-timing-of-ovulation.
us
an
43. Herbison AE. The gonadotropin-releasing hormone pulse generator.
Endocrinology. 2018;159(11):3723-3736.
M
initiation of the preovulatory LH surge in the human, Old World monkey and
e d
45. Abbara A, Eng PC, Phylactou M, et al. Kisspeptin receptor agonist has
2020;130(12):6739-6753.
Ac
46. Kenealy BP, Keen KL, Garcia JP, Kohlenberg LK, Terasawa E. Obligatory role
2017;114(52):13804-13809.
23
47. Marques P, Skorupskaite K, George JT, Anderson RA. Physiology of GNRH
and gonadotropin secretion. In: Feingold KR, Anawalt B, Boyce A, et al., eds.
t
ip
flutamide-metformin therapy reverses insulin resistance and reduces fat mass
cr
in nonobese adolescents with ovarian hyperandrogenism. J Clin Endocrinol
Metab. 2003;88(6):2600-2606.
us
50. Fleming R, Harborne L, MacLaughlin DT, et al. Metformin reduces serum
an
mullerian-inhibiting substance levels in women with polycystic ovary syndrome
51. Ortega MT, Carlson L, McGrath JA, et al. AMH is higher across the menstrual
Metab. 2020;105(4):e1762-e1771.
pt
women with polycystic ovary syndrome. Clin Exp Reprod Med. 2019;46(4):197-
ce
201.
Ac
53. Laven JS, Mulders AG, Visser JA, Themmen AP, De Jong FH, Fauser BC. Anti-
54. Lv PP, Jin M, Rao JP, et al. Role of anti-Müllerian hormone and testosterone in
24
55. Hendrix AO, Selgrade JF. Bifurcation analysis of a menstrual cycle model
Biol. 2014;361:31-40.
56. Vendola KA, Zhou J, Adesanya OO, Weil SJ, Bondy CA. Androgens stimulate
57. Andersen CY, Schmidt KT, Kristensen SG, Rosendahl M, Byskov AG, Ernst E.
t
ip
Concentrations of AMH and inhibin-B in relation to follicular diameter in normal
cr
human small antral follicles. Hum Reprod. 2010;25(5):1282-1287.
us
hormone in patients with polycystic ovary syndrome: relationship to the ovarian
an
follicle excess and to the follicular arrest. J Clin Endocrinol Metab.
2003;88(12):5957-5962.
M
59. Grossman MP, Nakajima ST, Fallat ME, Siow Y. Müllerian-inhibiting substance
60. Cimino I, Casoni F, Liu X, et al. Novel role for anti-Mullerian hormone in the
2016;7:10055.
Ac
61. Tata B, Mimouni NEH, Barbotin AL, et al. Elevated prenatal anti-Mullerian
62. Ruth KS, Day FR, Tyrrell J, et al. Using human genetics to understand the
258.
25
63. Pielecka J, Quaynor SD, Moenter SM. Androgens increase gonadotropin-
64. Silva MS, Prescott M, Campbell RE. Ontogeny and reversal of brain circuit
t
ip
cells. Hum Reprod. 2014;29(12):2736-2746.
cr
66. Qi X, Salem M, Zhou W, et al. Neurokinin B exerts direct effects on the ovary to
us
67. Blasco V, Pinto FM, Fernandez-Atucha A, et al. Altered expression of the
an
kisspeptin/KISS1R and neurokinin B/NK3R systems in mural granulosa and
Genet. 2019;36(1):113-120.
2020;114(4):869-878.
Ac
69. Yang F, Ruan YC, Yang YJ, et al. Follicular hyperandrogenism downregulates
Reproduction. 2015;150(4):289-296.
70. Barry JA, Azizia MM, Hardiman PJ. Risk of endometrial, ovarian and breast
26
LEGENDS
Figure 1. Patient disposition. aPatient experienced depressed mood, mood swings, headache, and
decreased libido considered by the investigator to be possibly related to treatment.
Figure 2. (A) Adjusted mean (SE) percentage change from baseline in total testosterone levels during
treatment with fezolinetant vs placebo. (B) Adjusted mean (SE) change from baseline in LH during
treatment with fezolinetant vs placebo. (C) Adjusted mean (SE) change from baseline in FSH during
treatment with fezolinetant vs placebo. (D) Effects of fezolinetant on LH-to-FSH ratio. All ITT
t
absolute LH-to-FSH ratios.
ip
FSH, follicle-stimulating hormone; ITT, intent-to-treat; LH, luteinizing hormone.
cr
Figure 3. Effects of fezolinetant on (A) adjusted mean change in E2 based on ANCOVA, (B)
endometrial thickness over time, and (C) menses frequency, ITT population.
a
us
P<0.05; bP<0.01. Change in E2 is based on least squares mean percentage change from the ANCOVA
model with treatment group as a fixed factor and baseline value as a covariate.
an
ANCOVA, analysis of covariance; E2, estradiol; ITT, intent-to-treat.
Figure 4. Effect of fezolinetant on (A) Anti-Müllerian hormone (AMH) and (B) adjusted mean change
M
27
Table 1. Demographics and baseline clinical characteristics
t
native
ip
Not askeda 4 (14.8) 5 (21.7) 5 (21.7)
Ethnicity, n (%)
cr
Hispanic/Latino 0 (0) 1 (4.3) 0 (0)
Not Hispanic/Latino 26 (96.3) 21 (91.3) 22 (95.7)
Not askeda 1 (3.7) 1 (4.3) 1 (4.3)
us
Abbreviations: BMI, body mass index.
a
Local regulations restricted asking.
an
M
e d
pt
ce
Ac
28
t
ip
cr
Table 2. Effect of fezolinetant on sex hormones (primary and secondary endpoints), intent-to-treat population
an
testosterone, Fezolinetant 60 mg 1.65 (0.66) 1.38 (0.59) −0.39 (0.12) 0.0379
nmol/L Fezolinetant 180 mg 2.16 (1.01) 1.39 (0.60) −0.80 (0.13) <0.0001
LH, IU/L Placebo 14.43 (6.51) 12.51 (6.62) −3.16 (1.04) —
Fezolinetant 60 mg 17.64 (16.83) 7.84 (4.31) −8.21 (1.18) 0.0022
M
Fezolinetant 180 mg 14.43 (8.60) 5.72 (4.47) −10.17 (1.28) <0.0001
FSH, IU/L Placebo 5.95 (1.86) 5.53 (1.68) −0.57 (0.26) —
Fezolinetant 60 mg 6.61 (4.41) 5.29 (1.40) −0.92 (0.30) 0.3770
ed
Fezolinetant 180 mg 5.67 (2.35) 4.65 (1.14) −1.46 (0.32) 0.0336
LH-to-FSH ratio Placebo 2.60 (1.32) 2.33 (1.10) −0.31 (0.16) —
Fezolinetant 60 mg 2.67 (1.62) 1.49 (0.77) −1.24 (0.18) 0.0003
pt
Fezolinetant 180 mg 2.67 (1.30) 1.16 (0.73) −1.45 (0.19) <0.0001
P4, ng/mL Placebo 1.31 (3.19) 2.12 (5.60) 0.42 (0.73) —
Fezolinetant 60 mg 2.56 (4.03) 1.00 (1.74) −0.77 (0.83) 0.2876
ce
an
Baseline 4.57 (0.25) 4.57 (0.27) 4.78 (0.28)
Change from baseline at week 12 −0.1 (0.17) 0 (0.19) −0.6 (0.29)
Body hair
M
Baseline 3.97 (0.36) 3.76 (0.39) 4.06 (0.45)
Change from baseline at week 12 −0.3 (0.17) −0.1 (0.17) −0.3 (0.26)
Weight ed
Baseline 3.92 (0.42) 3.74 (0.45) 4.10 (0.41)
Change from baseline at week 12 −0.3 (0.22) −0.4 (0.19) 0 (0.27)
a
Infertility Problems
pt
Baseline 4.70 (0.36) 4.37 (0.35) 4.77 (0.39)
Change from baseline at week 12 0.1 (0.24) 0.1 (0.26) −0.6 (0.35)
ce
Menstrual Problems
Baseline 4.38 (0.28) 3.67 (0.30) 4.14 (0.28)
Change from baseline at week 12 0.3 (0.26) 0.6 (0.26) 0.3 (0.30)
Ac
an
Baseline 16.9 (11.2) 14.4 (6.9) 17.8 (7.8)
Week 6 change −2.0 (8.7) 0.1 (7.3) −0.4 (11.2)
Week 12 change −0.7 (10.1) −1.8 (8.2) 3.7 (13.1)
M
Number of follicles, right ovary
Baseline 19.1 (12.8) 13.9 (7.0) 18.7 (12.8)
Week 6 change ed −2.1 (11.4) −1.0 (4.2) −2.1 (8.6)
Week 12 change −1.2 (9.8) 0.6 (6.6) 3.1 (12.4)
Surface of dominant follicle, mm3
Baseline 102.8 (103.6) 220.0 (354.4) 140.4 (255.5)
pt
Week 6 change −54.7 (113.9) −145.8 (349.9) −19.8 (151.3)
Week 12 change −24.0 (153.6) −99.2 (311.9) 15.6 (43.7)
ce
31
t
ip
cr
Table 5. Fezolinetant plasma concentrations, safety population
an
a
Predose samples
Week 1 (baseline) 23 0 (0) 0 — 22 0 (0) 0 —
Week 9 (trough) 20 57.51 (28.04) 19.2 0–576.0 17 362.06 (121.84) 153.0 0–1815.0
b
Postdose (peak) samples
M
Week 3 22 423.35 (57.53) 395.5 13.5–1023.0 18 1371.20 (149.55) 1431.5 9.7–2266.0
Week 6 21 468.91 (43.94) 440.0 0–847.0 17 1433.59 (110.24) 1415.0 753.0–2401.0
Week 12 ed 21 417.24 (59.07) 424.0 0–927.0 17 1361.88 (153.53) 1299.0 0–2428.0
a
Blood samples for fezolinetant concentrations were taken before intake of study drug.
b
Blood samples for fezolinetant concentrations were taken 2–8 hours after intake of study drug.
pt
ce
Ac
32
Table 6. TEAEs occurring in ≥3 patients and effects on liver function, safety population.
t
Rash 1 (3.7) 0 (0) 3 (13.0)
ip
Nasopharyngitis 5 (18.5) 3 (13.0) 2 (8.7)
Fatigue 3 (11.1) 2 (8.7) 0 (0)
Nausea 1 (3.7) 3 (13.0) 0 (0)
cr
Treatment-emergent AST or
ALT ≥3 x ULNa 1 0 1
us
Abbreviations: ALT, alanine aminotransferase; AST, aspartate aminotransferase; TEAE, treatment-
emergent adverse event; ULN, upper limit of normal.
a an
Based on laboratory testing; all increases were transient and resolved spontaneously during treatment.
33
Downloaded from https://academic.oup.com/jcem/advance-article/doi/10.1210/clinem/dgab320/6277155 by guest on 17 May 2021
t
ip
cr
Figure 1
us
34
an
M
ed
pt
ce
Ac
Downloaded from https://academic.oup.com/jcem/advance-article/doi/10.1210/clinem/dgab320/6277155 by guest on 17 May 2021
t
ip
cr
Figure 2A
us
35
an
M
ed
pt
ce
Ac
Downloaded from https://academic.oup.com/jcem/advance-article/doi/10.1210/clinem/dgab320/6277155 by guest on 17 May 2021
t
ip
cr
Figure 2B
us
36
an
M
ed
pt
ce
Ac
Downloaded from https://academic.oup.com/jcem/advance-article/doi/10.1210/clinem/dgab320/6277155 by guest on 17 May 2021
t
ip
cr
Figure 2C
us
37
an
M
ed
pt
ce
Ac
Downloaded from https://academic.oup.com/jcem/advance-article/doi/10.1210/clinem/dgab320/6277155 by guest on 17 May 2021
t
ip
cr
Figure2D
us
38
an
M
ed
pt
ce
Ac
Downloaded from https://academic.oup.com/jcem/advance-article/doi/10.1210/clinem/dgab320/6277155 by guest on 17 May 2021
t
ip
cr
Figure 3A
us
39
an
M
ed
pt
ce
Ac
Downloaded from https://academic.oup.com/jcem/advance-article/doi/10.1210/clinem/dgab320/6277155 by guest on 17 May 2021
t
ip
cr
Figure 3B
us
40
an
M
ed
pt
ce
Ac
Downloaded from https://academic.oup.com/jcem/advance-article/doi/10.1210/clinem/dgab320/6277155 by guest on 17 May 2021
t
ip
cr
Figure 3C
us
41
an
M
ed
pt
ce
Ac
Downloaded from https://academic.oup.com/jcem/advance-article/doi/10.1210/clinem/dgab320/6277155 by guest on 17 May 2021
t
ip
cr
Figure 4A
us
42
an
M
ed
pt
ce
Ac
Downloaded from https://academic.oup.com/jcem/advance-article/doi/10.1210/clinem/dgab320/6277155 by guest on 17 May 2021
t
ip
cr
Figure 4B
us
43
an
M
ed
pt
ce
Ac