Upload
others
View
2
Download
0
Embed Size (px)
Citation preview
Neurokinin B Receptor Antagonism in Women with Polycystic Ovary
Syndrome: A Randomized, Placebo-Controlled Trial
Jyothis T George1, Rahul Kakkar2, Jayne Marshall3, Martin L Scott2, Richard D Finkelman4, Tony W
Ho5, Johannes Veldhuis6, Karolina Skorupskaite7, Richard A Anderson7, Stuart McIntosh8, Lorraine
Webber3
1Diabetes Trials Unit, Oxford Centre for Diabetes, Endocrinology and Metabolism, Churchill
Hospital, Headington, Oxford, UK;
2AstraZeneca, Waltham, MA, USA;
3AstraZeneca, Mereside, Alderley Park, Macclesfield, Cheshire, UK;
4Formerly, AstraZeneca, Wilmington, DE, USA; currently, Shire, Wayne, PA, USA;
5AstraZeneca, Gaithersburg MD, USA;
6Endocrine Research Unit, Mayo Clinic College of Medicine, Center for Translational Science
Activities, Rochester, MN, USA;
7MRC Centre for Reproductive Health, The Queen's Medical Research Institute, University of
Edinburgh, Edinburgh, UK;
8Formerly, AstraZeneca, Mereside, Alderley Park, Macclesfield, Cheshire, UK; currently, Metis
Medical, Wilmslow, Cheshire, UK
Abbreviated Title: Neurokinin B Receptor Antagonism in PCOS
Key terms: polycystic ovary syndrome, neurokinin B receptor antagonism, randomized clinical trial
Word count (excluding abstract, figure captions, and references): 3933 words
Number of figures and tables: 3 figures, 1 table
1
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
Corresponding author and person to whom reprint requests should be addressed:
Lorraine Webber, AstraZeneca, Mereside, Alderley Park, Macclesfield, SK10 4TG. Email:
Funding: The study described herein was sponsored by AstraZeneca. It was designed and developed
by AstraZeneca in collaboration with clinical academics acting as consultants. These consultants were
from the Universities of Oxford (Dr Jyothis T George), Edinburgh (Professor Richard A Anderson),
and Chicago (Professor David A Ehrmann). Quintiles (London, UK) was contracted by AstraZeneca
to collect the data. Rahul Kakkar, Jayne Marshall, Martin L Scott, Richard D Finkelman, Tony W Ho,
Stuart McIntosh, and Lorraine Webber are current or former employees of AstraZeneca and
contributed to the study design, data analysis, and development of the manuscript. The decision to
publish the results described herein was taken at the start of the study, regardless of what the findings
would be, and the manuscript was reviewed by AstraZeneca before submission. Jyothis T George and
Lorraine Webber developed the manuscript outline; all other authors then had input into the outline.
All authors contributed to further of the manuscript, and approved the submission of this final version.
Disclosure statement: RK, JM, MLS, TWH, and LW are all employees of AstraZeneca. RDF and
SM were full-time employees of AstraZeneca at the time of the conduct of the study. JTG is the
International Co-ordinating Investigator for this study and served as a consultant for AstraZeneca and
Takeda. He has received consulting, speaking, travel, and/or research support from Amylin,
AstraZeneca, Boehringer Ingelheim, Bristol-Myers Squibb, Eli Lilly, Merck Sharp & Dohme, Novo
Nordisk, and Sanofi. RAA has worked as a consultant for AstraZeneca. KS and JV have no conflicts
of interest to disclose.
ClinicalTrials.gov identifier: NCT01872078
2
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
Abstract
Context: Polycystic ovary syndrome (PCOS), the most common endocrinopathy in women, is
characterized by high levels of secretion of luteinizing hormone (LH) and testosterone. Currently,
there is no treatment licensed specifically for PCOS.
Objective: To investigate whether a targeted therapy would decrease LH pulse frequency in women
with PCOS, subsequently reducing serum LH and testosterone concentrations and thereby presenting
a novel therapeutic approach to the management of PCOS.
Design: Double-blind, double-dummy, placebo-controlled, phase 2 trial.
Settings: University hospitals and private clinical research centres.
Participants: Women with PCOS aged 18–45 years.
Intervention: AZD4901 (a specific neurokinin-3 [NK3] receptor antagonist) at a dose of 20, 40, or
80 mg/day or matching placebo for 28 days.
Main outcome measure: Change from baseline in the area under the LH serum concentration–time
curve over 8 hours (AUC) on day 7 relative to placebo.
Results: Of a total of 67 randomized patients, 65 were evaluable. On day 7, the following baseline-
adjusted changes relative to placebo were observed in patients receiving AZD4901 80 mg/day: (1) a
reduction of 52.0% (95% CI: 29.6–67.3%) in LH AUC; (2) a reduction of 28.7% (95% CI: 13.9–
40.9%) in total testosterone concentration; and (3) a reduction of 3.55 LH pulses/8 hours (95% CI:
2.0–5.1) (all nominal P < .05).
Conclusions: The NK3 receptor antagonist AZD4901 specifically reduced LH pulse frequency and
subsequently serum LH and testosterone concentrations, thus presenting NK3 receptor antagonism as
a potential approach to treating the central neuroendocrine pathophysiology of PCOS.
3
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
Secondary Abstract
Results from this phase 2 clinical trial demonstrate the potential for a selective neurokinin-3
receptor antagonist to target the neuroendocrine pathophysiology of luteinizing hormone
hypersecretion and hyperandrogenism in PCOS.
4
70
71
72
73
Introduction
Polycystic ovary syndrome (PCOS) is the most common endocrinopathy in women, and it affects
approximately 5–10% of women of reproductive age (1,2). Different consensus groups have
developed different definitions of PCOS. Depending on which definition is used, diagnosis is based
on the presence of some or all of the following: chronic anovulation, polycystic appearance of the
ovaries, and excessive testosterone secretion (hyperandrogenemia) or activity (hyperandrogenism) (3-
5). PCOS is associated with several clinical presentations, such as menstrual dysfunction, infertility,
hirsutism, acne, obesity, and metabolic syndrome (4,6). In the long term, women with PCOS also
have an increased risk of type 2 diabetes mellitus and potentially cardiovascular disease (6,7).
Treatment involves management of symptoms or chronic suppression of the hypothalamic–pituitary
axis using a number of treatment modalities including metformin, anti-androgens and exogenous sex
steroids, often off-label (6,8). There is an unmet need to develop a targeted, safe, and effective
treatment that addresses the underlying central endocrinopathy.
The pathophysiological mechanisms underpinning PCOS are multi-factorial, including
developmental, metabolic and genetic factors. Nevertheless, PCOS is associated with an increase in
luteinizing hormone (LH) pulse amplitude and pulse frequency, which is likely driven by increased
pulsatile secretion of gonadotropin-releasing hormone (GnRH) (9). This excess of pituitary LH
secretion results in failure of ovulation and increased ovarian testosterone production (9). Recent
discoveries suggest that the kisspeptin-neurokinin B (NKB)–GnRH pathway is the pivotal regulator of
LH secretion (10,11). Indeed, patients with genetically impaired NKB signaling have low baseline LH
secretion and low LH pulse frequency (12,13). Thus, pharmacological NKB blockade may be a useful
approach to targeting the central pathophysiology of LH hypersecretion and hyperandrogenism in
PCOS.
In mammals, there are three tachykinin receptors, of which the neurokinin-3 (NK3) receptor appears
to be associated with a reproductive regulatory role through its ligand NKB (14). AZD4901 is a high-
affinity antagonist of the human NK3 receptor (15). It was initially developed for schizophrenia in
5
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
2007–2010 (as AZD2624) but did not meet its developmental efficacy goals for that indication (16).
In common with other NK3 receptor antagonists, however, AZD2624 reduced LH and testosterone
concentrations in healthy volunteers and patients without endocrine or reproductive disorders (15). At
the time, a reproductive role for NKB was yet to be elucidated (12). Since then, much evidence has
accrued that suggests that NKB has a central role in the generation of GnRH and thus LH pulsatility
(10). It is therefore thought that AZD4901 regulates pituitary LH and gonadal testosterone via
modulation of GnRH pulsatility.
PCOS is a heterogeneous disorder, with multiple pathophysiological mechanisms (e.g. insulin
resistance) in addition to LH hypersecretion contributing to its development. In this randomized
controlled trial, our intervention (AZD4901) specifically targets LH hypersecretion. We hypothesized
that AZD4901 could reduce LH pulsatility and prevent LH and possibly testosterone hypersecretion in
women with PCOS, and we investigated this hypothesis in a randomized, multicenter clinical trial.
Materials and Methods
Study design and participants
This randomized, double-blind, double-dummy, placebo-controlled, phase 2 trial (ClinicalTrials.gov
identifier: NCT01872078) was conducted between June 2013 and October 2014. Patients were
competitively recruited in nine centers in Germany, UK, and USA (Appendix 1). The study protocol
was reviewed and approved by the Institutional Review Board and Ethics Committee governing each
participating center, and the study was conducted in accordance with the Declaration of Helsinki.
Eligible patients were women aged 18–45 years with a body mass index of 18–40 kg/m2 and a clinical
diagnosis of PCOS; it was also a requirement that any confounding diagnosis had been excluded by
the investigator. Participants needed to meet of all of the following criteria: (1) polycystic ovaries
documented by ultrasound; (2) free testosterone >85% of the upper limit of reference range (measured
within 21 days prior to randomization at Arup Laboratories, US. Reference range (pg/mL): women
18-30 years 0.8 - 7.4; 31-40 years 1.3-9.2; 41-51 years 1.1-5.8); and (3) amenorrhea or
oligomenorrhea (defined as 6 menses per year).
6
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
Women who were not permanently or surgically sterile were required to use effective non-hormonal
methods of birth control, such as strict abstinence or use of effective non-hormonal methods of birth
control by the participant or their partner for the duration of the study. Acceptable barrier methods of
contraception included condom or occlusive cap (diaphragm or cervical/vault caps) with spermicidal
foam/gel/film/cream/suppository.
Patients were excluded if they had total testosterone serum concentrations ≥5 nmol/L (as very high
testosterone is often associated with alternative diagnoses such an androgen-secreting tumours), if
serum follicle-stimulating hormone [FSH] >10 IU/L (as a marker to exclude peri- or postmenopause),
or had menstruated within the last 30 days.
Women with uncontrolled hypertension/diabetes, or significant pulmonary, renal, hepatic, endocrine,
or other systemic disease, or any other clinically relevant diseases or abnormalities as judged by the
investigator were also excluded in this early phase clinical trial of an investigational medicinal
product. In addition, pregnant women and those not using adequate non-hormonal contraception were
excluded. Full exclusion criteria are presented in Supplemental Table 1.
Randomization and masking
Participants were randomized equally to four treatment groups: AZD4901 20 mg once daily
(20 mg/day), AZD4901 20 mg twice daily (40 mg/day), AZD4901 40 mg twice daily (80 mg/day), or
placebo twice daily (Fig. 1). These doses were selected based on data from previous dose escalation
studies, including those in which ascending doses of AZD4901 up to 80 mg/day were administered to
healthy volunteers. In these studies, significant LH and testosterone suppression was seen at 40
mg/day, allowing a range of safe doses to be included in this study to explore a dose-response
relationship.
Sequential randomization was carried out in each study center by the investigator following a blinded
computer-generated randomization scheme produced by Quintiles Early Clinical Development
(London, UK) using the AstraZeneca Global Randomization system. To maintain study blinding,
AZD4901 and/or matching placebo were administered such that two tablets were taken twice daily by
7
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
all participants. The first dose of AZD4901 or placebo was administered on the morning of day 1.
Participants were treated for 28 days.
Procedures and outcomes
The primary endpoint was the change from baseline in the area under the LH plasma concentration–
time curve over 8 hours post-dose (AUC) on day 7 relative to placebo. Day 7 was selected as the time
point to evaluate LH because AZD4901 would have achieved steady state and because any
confounding that may occur owing to a LH surge preceding spontaneous ovulation would be avoided.
Women who had menstruated in the 30 days before the baseline visit (during which screening
procedures and lab tests were undertaken) were excluded.
Secondary objectives were to evaluate: (1) the change from baseline in average total and free
testosterone serum concentrations over 8 hours post-dose (Cavg) on days 7 and 28 relative to placebo;
(2) the safety and tolerability of AZD4901; (3) the pharmacokinetics of AZD4901 and its major
metabolite AZD12292232; and (4) the pharmacokinetic/pharmacodynamic effect of AZD4901 on LH
and testosterone concentrations and on LH pulsatility parameters on days 7 and 28 relative to placebo.
Several exploratory endpoints were investigated: (1) the change from baseline in LH AUC on day 28
relative to placebo; (2) changes from baseline in FSH, estradiol, progesterone, prolactin, thyroid-
stimulating hormone, and insulin-like growth factor-1 on days 7 and 28; (3) glycated hemoglobin
concentration on day 28; (4) the impact of AZD4901 on health-related quality of life from baseline to
day 28; and (5) the impact of AZD4901 on PCOS-specific patient-reported outcomes as
measured by changes from baseline on days 7, 14, 21, and 28.
In addition, two post hoc exploratory analyses were carried out to assess: (1) the absolute change from
baseline in LH AUC:FSH AUC ratio at days 7 and 28 relative to placebo; and (2) the changes in LH
AUC, total and free testosterone Cavg, and LH pulsatility parameters relative to placebo in patients
with no biochemical evidence of ovulation (serum progesterone <6 ng/dL [19.1 nmol/L] at all study
visits). Patients with no biochemical evidence of ovulation are referred to as ‘non-ovulating patients’
hereafter.
8
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
LH pulsatility assessments were carried out using peripheral venous blood samples collected at
baseline and on days 7 and 28 at 10-minute intervals for 8 hours on these three days. FSH and total
and free testosterone concentrations were assessed using samples collected at baseline and on days 7
and 28 before the morning dose and then every hour for 8 hours. Estradiol, progesterone, prolactin,
thyroid-stimulating hormone, insulin-like growth factor, total and free thyroxin, and glycated
hemoglobin were measured using single samples collected at baseline and on days 7 and 28. Analyses
of total and free testosterone were performed using high-performance liquid chromatography tandem
mass spectrometry. All other endocrine markers were analyzed immunometrically.
Samples for pharmacokinetic analysis were collected on days 7 and 28 before the morning dose and at
20 minutes, 40 minutes, and 1, 1.5, 2, 3, 4, 6, and 8 hours post-dose. Health-related quality of life was
assessed using the 36-item Short-Form Health Survey (SF-36), completed by patients at baseline and
on day 28. Safety assessments included adverse event monitoring, vital sign measurements,
electrocardiograms, and physical examination. In addition, the Columbia-Suicide Severity Rating
Scale (C-SSRS) was used to identify any suicide-related adverse events, including suicidal behaviour
and ideation, and was administered at baseline and each visit throughout the study. The C-SSRS was
included because it had previously been mandated by the US Food and Drug Administration for the
early clinical programme of AZD4901 for the indication of schizophrenia.
Statistical methods
It was determined that a sample size of 48 patients (12 patients per treatment group) would be
required to detect a 30% change from baseline in LH AUC on day 7 relative to placebo with 76%
power at the two-sided 5% significance level. Allowing for potential drop-outs, it was planned to
randomize 56 patients to achieve 12 evaluable patients per group. A total of 67 patients were actually
randomized, 65 of whom were evaluable.
The dataset for analysis of pharmacodynamic parameters included all patients who received at least
one dose of study medication (AZD4901 or placebo) and had appropriate pharmacodynamic
measurement. The safety dataset comprised all patients who received at least one dose of study
9
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
medication and for whom some post-dose data were available. The pharmacokinetic analysis set
comprised patients who received at least one dose of AZD4901 and had at least one post-dose
pharmacokinetic measurement without important protocol deviations or violations that could have
affected the pharmacokinetic parameters significantly.
All analyses were performed using a mixed-effects model for repeated measures (MMRM) on the ln-
transformed ratio to baseline, with repeated-effects for day, fixed-effects for treatment, and treatment-
by-day interaction. No adjustments were made for multiplicity.
LH AUC was calculated by linear up/linear down trapezoidal summation of observed serum
concentrations. Data from day –1, 7, and 28 samples collected outside a ±2-minute collection window
were not included in descriptive statistics for LH by time point. LH AUC was calculated if there were
no more than five non-consecutive missing values in the profile and no more than three consecutive
missing values (no more than two consecutive missing values if one was at 0 or 8 hours). For LH
AUC, comparisons between AZD4901 and placebo were performed using a MMRM on the ln-
transformed ratio to baseline, with ln-transformed baseline LH AUC included as a covariate.
Data from day –1, 7, and 28 samples collected outside a ±10-minute collection window were not
included in descriptive statistics for total and free testosterone serum concentrations by time. Total
and free testosterone Cavg were calculated if there were no more than two consecutive or non-
consecutive missing values in the profile. If a value at 0 or 8 hours was missing, total testosterone C avg
was imputed using the next or previously scheduled value; free testosterone Cavg was not calculated if
a value at 0 or 8 hours was missing. For total and free testosterone, comparisons between AZD4901
and placebo were performed using a MMRM on the ln-transformed ratio to baseline, with ln-
transformed baseline Cavg included as a covariate.
Pharmacokinetic parameters were derived using standard non-compartmental methods with
WinNonlin Professional version 6.3 (Pharsight Corp., Mountain View, CA, USA) and descriptive
statistics were reported. Descriptive statistics were also reported for health-related quality of life
10
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
individual scale scores and for adverse events. Statistical analyses were performed using SAS version
9.4 (SAS Institute, Cary, NC, USA).
Where the analysis has been performed using ln transformed data (i.e. LH AUC), the descriptive
measures presented in the paper relate to the geometric mean and where the analysis has been
performed using non transformed data (i.e. Number of pulses per 8 hours), descriptive measures
presented relate to the arithmetic mean.
LH pulsatility deconvolution analysis
The number of LH pulses, the mass-per-pulse (MPP), and LH basal secretion in 8 hours were derived
using deconvolution analyses described previously (17,18). Deconvolution estimates were calculated
if there were no more than three non-consecutive missing values in the profile and no more than two
consecutive missing values in the profile. A MMRM on the ln-transformed ratio to baseline values
was used for comparisons between AZD4901 and placebo for MMP and LH basal secretion, with the
ln-transformed baseline values included as a covariate. For the number of pulses, comparisons
between AZD4901 and placebo were carried out using a MMRM on the absolute change from
baseline, with the baseline value included as a covariate.
Results
Study population
Of the 403 women assessed for eligibility, 67 met the inclusion criteria and were randomized to
treatment; two patients did not receive an intervention because of difficult venous access, leaving 65
evaluable patients (Fig. 1). The most common reason for non-eligibility was failure to meet the
screening criteria for free testosterone level (50% of screen fails); around one third of screen fails did
not meet other laboratory inclusion parameters (most frequently, minor abnormalities of ALT/AST
likely associated with mild steatohepatitis, iron-deficiency anaemia and rare cases of elevated HbAlc),
whilst the remainder failed various clinical criteria including menstruation within the last month and
body mass index. Of the 65 evaluable patients, 15 patients received AZD4901 20 mg/day, 17 received
11
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
AZD4901 40 mg/day, 17 received AZD4901 80 mg/day, and 16 received placebo. Demographic data
and baseline characteristics for these 65 patients are shown in Table 1.
Primary endpoint
The baseline-adjusted changes of LH AUC at day 7 for the AZD4901 groups relative to placebo are
presented in Fig. 2A. In the AZD4901 80 mg/day group, there was a baseline-adjusted reduction in
LH AUC of 52.0% (95% CI: 29.6–67.3%; P = .0003) relative to placebo. There was no evidence of
an effect on LH AUC change from baseline to day 7 relative to placebo for the lower AZD4901 doses.
Secondary endpoints
Change in serum testosterone concentration
In the AZD4901 80 mg/day group, there was a baseline-adjusted reduction in total testosterone Cavg of
28.7% (95% CI: 13.9–40.9%; P = .0006) on day 7 relative to placebo. A corresponding reduction in
free testosterone Cavg of 19.2% (95% CI: 0.14–34.62%; P = .0486) was also observed in this group on
day 7. Similarly to LH, no significant reductions in testosterone concentrations from baseline to day 7
were observed in the groups receiving the lower AZD4901 doses relative to the group receiving
placebo (Fig. 2C and Supplemental Table 2). There was no evidence of an effect of any AZD4901
dose on total and free testosterone concentrations at day 28 (Fig. 2C and Supplemental Table 2).
To explore the effect of AZD4901 on testosterone in women who were considered to have no
biochemical evidence of ovulation in the study, a post hoc exploratory analysis that excluded patients
with serum progesterone ≥6 ng/dL [19.1 nmol/L] at any study visit was carried out (23). Nine women
were considered to have ovulated during the study: three each in the AZD4901 20 mg/day and 80
mg/day groups, two in the AZD4901 40 mg/day group, and one in the placebo group. When these
women were excluded, the baseline-adjusted reductions in total testosterone Cavg from baseline to days
7 and 28 relative to placebo were 27.1% (95% CI: 13.3–38.7%) and 20.8% (95% CI: 5.3–33.8%),
respectively, in the AZD4901 80 mg/day group (Fig. 2D and Supplemental Table 2). Corresponding
reductions in free testosterone Cavg were 22.8% (95% CI: 6.8–36.0%) and 23.8% (95% CI: 7.3–
12
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
37.3%) (Supplemental Table 2). The baseline characteristics of non-ovulating patients were not
numerically different from that of the whole cohort.
Luteinizing hormone pulsatility parameters
LH pulse frequency and LH basal secretion were significantly reduced in the AZD4901 80 mg/day
group on day 7 relative to placebo. There was a greater decrease in the number of LH pulses from
baseline to day 7 in the AZD4901 80 mg/day group than in the placebo group, with the difference
being 3.55 (95% CI: 2.0–5.1) pulses/8 hours (Fig. 3A and Supplemental Table 2). The reduction in
LH basal secretion from baseline to day 7 relative to placebo was 78.8% (95% CI: 53.6–90.3%) (Fig.
3C and Supplemental Table 2); mass-per-pulse remained unchanged (Fig. 3E and Supplemental Table
2). These effects persisted in non-ovulating patients (Fig. 3B and D and Supplemental Table 2).
Safety and tolerability
There was one serious adverse event reported in the study: a case of appendicitis considered by the
investigator to be unrelated to treatment but which led to study discontinuation. This was the only
adverse event that led to discontinuation. Overall, adverse events were reported by 32 out of 49
patients receiving AZD4901 (65.3%) and 8 out of 16 patients receiving placebo (50.0%). The most
common preferred terms for adverse events reported by patients were headache (reported by 14
patients [21.5%]; six assessed by the investigators to be related to treatment), nasopharyngitis
(reported by five patients [7.7%]; none assessed by the investigators to be related to treatment) and
dizziness (reported by three patients [4.6%]; one assessed by the investigators to be related to
treatment). No dose dependency was discernible in this small number of events (Supplemental Table
5).
No patients reported suicidal ideation or behaviour on the C-SSRS questionnaire
while receiving treatment or during follow-up.
Pharmacokinetic endpoints
Circulating concentrations of AZD4901 and its active metabolite increased in a dose-dependent
manner that was consistent with the previously reported pharmacokinetic profile (16), and steady state
13
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
was reached by day 7 (Supplemental Table 3). AZD4901 was quickly absorbed following
oral dosing; time to maximum concentration was approximately 1.5–2 hours for all doses on days
7 and 28.
Exploratory endpoints
There was no evidence of an effect of AZD4901 on changes in LH AUC from baseline to day 28
relative to placebo (Fig. 2A and Supplemental Table 2). In the post hoc analysis, when considering
only non-ovulating patients, the change in LH AUC from baseline to day 28 in the AZD4901 80
mg/day group was 34.9% (95% CI: 6.6–54.6%) relative to placebo (Fig. 2B and Supplemental Table
2).
Relevant biochemical parameters are summarized in Supplemental Table 4. FSH concentrations
remained largely unchanged in all treatment groups. Therefore, given the changes in LH AUC, there
was evidence of an absolute reduction in the baseline-adjusted LH AUC:FSH AUC ratio relative to
placebo. Reductions relative to placebo were observed on day 7 (0.70; 95% CI: 0.23–1.17) and day 28
(0.72; 95% CI: 0.23–1.21) in the AZD4901 80 mg/day group. There was no evidence that the lower
AZD4901 doses had an effect on the LH AUC:FSH AUC ratio relative placebo (Supplemental Table
2).
There was no evidence of an effect of AZD4901 on health-related quality of life. Changes from
baseline to day 28 across the seven parameters of the SF-36 questionnaire were small across the four
treatment groups, and there were no obvious trends (Supplemental Table 6).
Discussion
In this first study to manipulate the NKB–GnRH pathway in PCOS, the NK3 receptor antagonist
AZD4901 specifically reduced LH pulse frequency and, subsequently, serum LH and testosterone
concentrations. These reductions persisted in non-ovulating patients until the end of the dosing period
(day 28), although were not statistically significant in the whole group. Longer studies assessing
clinical outcomes (e.g. ovulation and hirsutism) and quantification of potential metabolic
14
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
improvements in larger populations, as well as potential compensatory mechanisms, are
needed to explore further this potential therapeutic approach.
The present results are consistent with the hypothesis that modulation of the GnRH axis by NK3
antagonism using AZD4901 would decrease LH pulse frequency and lower LH and testosterone
concentrations in women with PCOS. They are also consistent with previous data from early clinical
investigations of NK3 antagonists in individuals without an endocrine or reproductive disorder. In
these studies in healthy volunteers and patients with schizophrenia, dose-dependent decrease in LH
and testosterone concentrations were observed (15). Taken together with a central role of NKB in the
regulation of GnRH and LH pulse frequency, the present study demonstrates the potential of NKB
antagonism to provide a novel approach to treating the central neuroendocrine pathophysiology of
PCOS (i.e. the LH hypersecretion that, in turn, drives androgen excess).
In this study, reduction in overall LH secretion was underpinned by reductions in LH pulsatility as
well as in basal LH secretion but not in the amount of LH secreted per pulse. These findings are
consistent with low LH pulsatility observed in patients with genetic defects leading to impaired NKB
signaling (13). Therefore, the present study contributes to the recent insights obtained into the
regulation of GnRH pulsatility following discoveries of hypothalamic roles for kisspeptin and
neurokinin B (10,19,20).
The observed reductions in LH and testosterone from baseline seen with the highest dose of AZD4901
were statistically significant at day 7 for the study population, but not at day 28. Reductions were,
however, statistically significant at day 28 in those women who did not ovulate during the study. This
is because, given the small sample size, the day 28 LH and testosterone results were confounded by
what appeared to be a pre-ovulatory LH surge in a small number of women; excluding women who
had ovulated during the study from the analysis resulted in the maintenance of the reduction in LH
and testosterone to day 28.
AZD4901 was well tolerated in patients with PCOS. Most adverse events were considered unrelated
to the study medication by the investigators, including the single serious adverse event. Furthermore,
15
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
given this class of compounds was initially developed for the treatment of schizophrenia, it is
reassuring to note that questionnaires assessing participant well-being (SF-36 and C-SSRS) did not
show any signals of concern.
Currently, while several medications are used to treat PCOS and its symptoms, there is no treatment
with specific regulatory approval for PCOS, and there are very few new molecular entities in clinical
development for this condition (8). Hence, a range of agents such as spironolactone, GnRH
modulators, metformin, oral contraceptive pills and clomiphene are used to manage the symptoms and
associated health complications of PCOS (6), reflecting the multifactorial aetiology of the condition.
The results from this study, if consistently reproduced in subsequent clinical studies, suggest that
AZD4901 has the potential to emerge as a novel therapy for PCOS and to complement recent
developments in the treatment of anovulatory infertility in women with PCOS (22).
The present study has clear strengths such as the inclusion of detailed LH pulse profiling and the use
of placebo control to estimate placebo-adjusted changes from baseline hormones. Only the highest
dose administered elicited a significant response in our study, suggesting that we may not have
reached maximal response in this PCOS population and that testing doses higher than 80 mg/day in
future studies may be warranted, a point supported by the safety and tolerability profile of AZD4901
in the present study. Before AZD4901 can be developed as a therapy for PCOS, longer studies
assessing clinical outcomes (e.g. ovulation and hirsutism) and quantification of potential metabolic
improvements in larger populations, as well as potential compensatory mechanisms, are needed.
Because this was a phase 2a trial aimed at validating the concept, our focus was on biomarkers such
as LH and testosterone; the duration of treatment was insufficient to assess the effects on clinical
endpoints such as ovulation. A small number of patients appear to have ovulated during the trial,
based on random serum progesterone >6 ng/mL (19.1 nmol/L), which is consistent with clinical
practice recommendations (23) and previous data (24). A total of nine women ovulated, three each in
the 20 mg/day and 80 mg/day groups, two in the 40 mg/day group, and one in the placebo group.
These small numbers did not allow us to make any meaningful comparisons between groups, or
16
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
between ovulators and non-ovulators. While the observed ovulation rates suggested by elevation of
serum progesterone (particularly among treated patients) may be higher during our study than
expected for PCOS patients in general, their possible relationship to treatment cannot be firmly
concluded for at least two reasons: firstly, the small numbers of ovulating patients observed do not
allow statistically rigorous comparisons among groups; secondly, the timing of ovulation within the
study appeared to differ across the nine women but serum progesterone measurements were only
taken at baseline and days 7, 28 and 42, so the day of ovulation cannot be precisely identified. Both
reasons reflect the fact that this early study was designed and powered to achieve a different primary
endpoint. Given the clinical importance of menstrual irregularity in PCOS, ovulation needs to be
characterized further in future longer-term studies using self-reported menstruation (e.g. menstrual
diary), biomarkers (e.g. LH, estradiol), and/or ultrasonography over multiple cycles.
The results of our study also have implications for wider research into new therapies for patients with
PCOS. First, the heterogeneity of the PCOS phenotype presents a challenge to attaining adequate
power in early-phase randomized controlled trials. We addressed this by selecting a
hyperandrogenemic population with polycystic ovarian morphology and menstrual irregularity. Such
an approach, however, required the screening of well over 400 women to recruit 67 participants.
Furthermore, the generalizability of our results to non-hyperandrogenemic patients with PCOS
requires further study.
Finally, it has to be emphasized that the present study is a clinical trial of a pharmacological agent;
therefore, inferences on the aetiology of PCOS and the multi-factorial nature of the mechanism by
which LH pulse frequency becomes increased cannot be drawn from the present data.
In conclusion, this is the first clinical study to manipulate the hypothalamic kisspeptin-NKB–GnRH
pathway in women with PCOS. The NK3 receptor antagonist AZD4901 reduced serum LH pulse
frequency and, subsequently, serum LH and testosterone concentrations. These findings demonstrate
the potential for NKB antagonism to provide a novel therapeutic approach by targeting the
neuroendocrine pathophysiology in PCOS.
17
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
Acknowledgments
We would like to thank Professor David A Ehrmann, Professor Rury R Holman, Professor Richard S
Legro, Professor John C Marshall, Professor Robert P Millar, and Dr Stephanie Seminara for their
input on the design of this study and/or for their thoughtful comments on this manuscript. We are also
grateful to Chris Davison (AstraZeneca) for providing additional statistical input. Medical writing
support was provided by Stéphane Pintat, PhD, of Oxford PharmaGenesis, Oxford, UK, and was
funded by AstraZeneca.
18
408
409
410
411
412
413
414
References
1. March WA, Moore VM, Willson KJ, Phillips DI, Norman RJ, Davies MJ. The
prevalence of polycystic ovary syndrome in a community sample assessed under
contrasting diagnostic criteria. Hum Reprod. 2010; 25:544–551.
2. Norman RJ, Dewailly D, Legro RS, Hickey TE. Polycystic ovary syndrome.
Lancet. 2007; 370:685–697.
3. Azziz R, Carmina E, Dewailly D, Diamanti-Kandarakis E, Escobar-Morreale
HF, Futterweit W, Janssen OE, Legro RS, Norman RJ, Taylor AE, Witchel SF.
Positions statement: criteria for defining polycystic ovary syndrome as a
predominantly hyperandrogenic syndrome: an Androgen Excess Society guideline. J
Clin Endocrinol Metab. 2006; 91:4237–4245.
4. Legro RS, Arslanian SA, Ehrmann DA, Hoeger KM, Murad MH, Pasquali R,
Welt CK. Diagnosis and treatment of polycystic ovary syndrome: an Endocrine
Society clinical practice guideline. J Clin Endocrinol Metab. 2013; 98:4565–4592.
5. Rotterdam ESHRE/ASRM-Sponosred PCOS Consensus Workshop Group.
Revised 2003 consensus on diagnostic criteria and long-term health risks related to
polycystic ovary syndrome. Fertil Steril. 2004; 81:19–25.
6. Azziz R, Marin C, Hoq L, Badamgarav E, Song P. Health care-related economic
burden of the polycystic ovary syndrome during the reproductive life span. J Clin
Endocrinol Metab. 2005; 90:4650–4658.
7. Welt CK, Carmina E. Clinical review: Lifecycle of polycystic ovary syndrome
(PCOS): from in utero to menopause. J Clin Endocrinol Metab. 2013; 98:4629–4638.
8. Rocca ML, Venturella R, Mocciaro R, Di Cello A, Sacchinelli A, Russo V,
Trapasso S, Zullo F, Morelli M. Polycystic ovary syndrome: chemical
pharmacotherapy. Expert Opin Pharmacother. 2015; 16:1369–1393.
19
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
9. Marshall JC, Eagleson CA, McCartney CR. Hypothalamic dysfunction. Mol Cell
Endocrinol. 2001; 183:29–32.
10. Skorupskaite K, George JT, Anderson RA. The kisspeptin-GnRH pathway in
human reproductive health and disease. Hum Reprod Update. 2014; 20:485–500.
11. George JT, Seminara SB. Kisspeptin and the hypothalamic control of reproduction:
lessons from the human. Endocrinology. 2012; 153:5130–5136.
12. Topaloglu AK, Semple RK. Neurokinin B signalling in the human reproductive axis.
Mol Cell Endocrinol. 2011; 346:57–64.
13. Young J, George JT, Tello JA, Francou B, Bouligand J, Guiochon-Mantel A,
Brailly-Tabard S, Anderson RA, Millar RP. Kisspeptin restores pulsatile LH
secretion in patients with neurokinin B signaling deficiencies: physiological,
pathophysiological and therapeutic implications. Neuroendocrinology. 2013; 97:193-
202.
14. Maeda K, Ohkura S, Uenoyama Y, Wakabayashi Y, Oka Y, Tsukamura H,
Okamura H. Neurobiological mechanisms underlying GnRH pulse generation by the
hypothalamus. Brain research. 2010; 1364:103–115.
15. Malherbe P, Ballard TM, Ratni H. Tachykinin neurokinin 3 receptor antagonists: a
patent review (2005–2010). Expert Opin Ther Pat. 2011; 21:637–655.
16. Litman RE, Smith MA, Desai DG, Simpson T, Sweitzer D, Kanes SJ. The
selective neurokinin 3 antagonist AZD2624 does not improve symptoms or cognition
in schizophrenia: a proof-of-principle study. J Clin Psychopharmacol. 2014; 34:199–
204.
17. Liu PY, Keenan DM, Kok P, Padmanabhan V, O'Byrne KT, Veldhuis JD.
Sensitivity and specificity of pulse detection using a new deconvolution method. Am
J Physiol Endocrinol Metab. 2009; 297:E538–544.
20
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
18. Veldhuis JD, Keenan DM, Pincus SM. Motivations and methods for analyzing
pulsatile hormone secretion. Endocr Rev. 2008; 29:823–864.
19. Herbison AE. Physiology of the adult GnRH neuronal network. In: Plant TM,
Zeleznik AJ, eds. Knobil and Neill's Physiology of Reproduction. San Diego:
Academic Press; 2014:399–467.
20. Jayasena CN, Comninos AN, De Silva A, Abbara A, Veldhuis JD, Nijher GM,
Ganiyu-Dada Z, Vaal M, Stamp G, Ghatei MA, Bloom SR, Dhillo WS. Effects of
neurokinin B administration on reproductive hormone secretion in healthy men and
women. J Clin Endocrinol Metab. 2014; 99:E19–27.
21. Sirmans SM, Pate KA. Epidemiology, diagnosis, and management of polycystic
ovary syndrome. Clin Epidemiol. 2013; 6:1–13.
22. Legro RS, Brzyski RG, Diamond MP, Coutifaris C, Schlaff WD, Casson P,
Christman GM, Huang H, Yan Q, Alvero R, Haisenleder DJ, Barnhart KT,
Bates GW, Usadi R, Lucidi S, Baker V, Trussell JC, Krawetz SA, Snyder P, Ohl
D, Santoro N, Eisenberg E, Zhang H. Letrozole versus clomiphene for infertility in
the polycystic ovary syndrome. N Engl J Med. 2014; 371:119–129.
23. Welt CK. Evaluation of the menstrual cycle and timing of ovulation. Waltham, MA:
Uptodate; 2014: http://www.uptodate.com/contents/evaluation-of-the-menstrual-
cycle-and-timing-of-ovulation. Accessed December 2014
24. Leiva R, Bouchard T, Boehringer H, Abulla S, Ecochard R. Random serum
progesterone threshold to confirm ovulation. Steroids. 2015; 101:125-129.
21
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
Table 1. Patient demographics and baseline characteristics
AZD4901 20 mg/day AZD4901 40 mg/day AZD4901 80 mg/day Placebo
(n = 15) (n= 17) (n = 17) (n = 16)
Age, yearsa 29 (6) 27 (6) 28 (6) 27 (3)
Height, cma 165.4 (6.2) 164.5 (8.1) 161.7 (4.5) 165.9 (7.4)
Weight, kga 85.8 (16.9) 84.2 (17.0) 85.2 (18.6) 87.9 (20.2)
BMI, kg/m2,a 31.1 (5.9) 30.8 (5.6) 32.2 (6.2) 31.9 (6.6)
Race, n (%)
White 15 (100.0) 13 (76.5) 11 (64.7) 14 (87.5)
Black or African American 0 (0.0) 3 (17.6) 3 (17.6) 1 (6.3)
Asian 0 (0.0) 0 (0.0) 2 (11.8) 0 (0.0)
Other 0 (0.0) 1 (5.9) 1 (5.9) 1 (6.3)
Ethnicity, n (%)
Hispanic 2 (13.3) 2 (11.8) 1 (5.9) 3 (18.8)
Non-Hispanic 13 (86.7) 15 (88.2) 16 (94.1) 13 (81.3)
Serum hormone Cavgb
LH, IU/L 9.78 (3.49) 9.12 (4.59) 9.22 (3.76) 9.09 (5.09)
FSH, IU/L 6.15 (2.06) 4.57 (1.67) 4.52 (1.49) 4.68 (1.35)
22
487
Total testosterone, nmol/L 1.96 (0.624) 2.07 (0.921) 2.25 (0.616) 1.68 (0.680)
Free testosterone, pmol/L 66.3 (37.8) 72.3 (32.7) 91.9 (30.1) 84.5 (57.4)
Data are arithmetic mean (standard deviation) unless otherwise stated.
aAssessed during screening.
bAssessed at baseline.
Cavg, average concentration over 8 hours; FSH, follicle-stimulating hormone; LH, luteinizing hormone.
23
488
489
490
491
492
Fig. 1. Patient disposition
Numbers of patients reported for days 7 and 28 are for changes in luteinizing hormone area under the concentration–time curve from baseline (0–8 hours
post-dose). Patients excluded from the analysis at day 7 because of an incomplete profile were not necessarily excluded from the analysis at day 28.
24
493
494
495
496
Fig. 2. Changes in LH AUC and total testosterone Cavg at days 7 and 28 relative to placebo. (A) and
(B) show baseline-adjusted changes in geometric means of LH AUC relative to placebo for all
analyzed patients and non-ovulating patients, respectively. (C) and (d) show baseline-adjusted
changes in geometric means of total testosterone Cavg relative to placebo for all analyzed patients and
non-ovulating patients, respectively. Non-ovulating patients were those with no biochemical evidence
of ovulation (serum progesterone <6 ng/dL [19.1 nmol/L] at all study visits). Whiskers represent 95%
confidence intervals.
AUC, area under the concentration–time curve (0–8 hours post-dose); Cavg, average concentration
over 8 hours post-dose; LH, luteinizing hormone. Dotted vertical lines on individual figures highlight
zero (i.e., no change from baseline).
25
497
498
499
500
501
502
503
504
505
506
507
Fig. 3. Changes in LH pulsatility parameters at days 7 and 28 relative to placebo.
(A) and (B) show changes in baseline-adjusted arithmetic means of number of LH pulses relative to
placebo for all analyzed patients and non-ovulating patients, respectively. (C) and (D) show baseline-
adjusted changes in geometric means of LH basal secretion relative to placebo for all analyzed
patients and non-ovulating patients, respectively. (E) and (F) show changes in baseline-adjusted
geometric means of LH MPP relative to placebo for all analyzed patients and non-ovulating patients,
respectively. Non-ovulating patients were those with no biochemical evidence of ovulation (serum
progesterone <6 ng/dL [19.1 nmol/L] at all study visits). Whiskers represent 95% confidence
intervals. LH, luteinizing hormone; MPP, mass-per-pulse. Dotted vertical lines on individual figures
highlight zero, i.e., no change from baseline.
26
508
509
510
511
512
513
514
515
516
517
518
Appendix 1. List of participating sites
1. Miami Research Associates, Miami, FL, USA
2. Bio-Kinetic Clinical Applications, Inc., Springfield, MO, USA
3. Washington University School of Medicine, Department of Obstetrics and Gynecology,
Division of Reproductive Endocrinology, St Louis, MO, USA
4. Charité Research Organisation GmbH, Berlin, Germany
5. BioKinetic Europe, Belfast, Northern Ireland
6. Edinburgh Fertility and Reproductive Endocrine Centre, Royal Infirmary of Edinburgh,
Edinburgh, UK
7. Quintiles Drug Research Unit at Guy’s Hospital, London UK
8. Compass Research Phase 1, Orlando, FL, USA
9. The University of Chicago, Department of Medicine, Section of Endocrinology, Chicago, IL,
USA
27
519
520
521
522
523
524
525
526
527
528
529
530
531
Supplemental Table 1. Exclusion criteria
Patients were excluded from the study if they met the following criteria
Were perimenopausal or had reached menopause (defined as FSH concentration >10 IU/L)
Menstruation in the 28 days before the baseline visit
Presence of a clinically relevant disease or abnormalities (particularly abnormal vaginal bleeding) that prevented the patient from
participating in the study, put them at risk, or would interfere with the study results
A significant illness in the 2 weeks preceding the study
Evidence of uncontrolled hypertension (defined as systolic blood pressure ≥160 mmHg and/or diastolic blood pressure ≥100 mmHg);
uncontrolled diabetes; or significant pulmonary, renal, hepatic, endocrine, or other systemic disease
A hysterectomy or bilateral oophorectomy or both
History of Gilbert's syndrome, infectious hepatitis, or other significant hepatic disease
History of gastric or small intestinal surgery or current disease that causes malabsorption
History of or current hyperthyroidism
Abnormal ECG, marked prolongation of QT/QTc interval, additional risk factors for Torsades de Pointes, or the use of concomitant
medications that prolong the QT/QTc interval
Positive human immunodeficiency virus (HIV), hepatitis B, or hepatitis C serology at screening
28
532
History of hypersensitivity to >2 chemical classes of drugs
Alcohol or substance abuse or consumed ≥3 alcoholic drinks per day
Blood loss of >200 mL in the 30 days from baseline, >500 mL in the 56 days from baseline, >1350 mL in 1 year from baseline, or donation
of blood products in the 14 days before baseline
Neoplastic disease in the past 5 years (except adequately treated basal cell, squamous cell skin cancer, or in situ cervical cancer)
Abnormal or unexplained laboratory test results (aspartate aminotransferase >1.5 times ULN; alanine aminotransferase >1.5 times ULN;
total bilirubin >1.5 times ULN; serum creatinine >2.0 times ULN; hematocrit less than LLN; prolactin >2.0 times ULN)
Withdrawal from oral contraceptives and LH concentrations <3 IU/L in the 7 days before the start of the study
Use of potent or moderate CYP3A4 or CYP2C9 inhibitors, potent or moderate CYP3A4 or CYP2C9 inducers, hormonal contraceptives,
antiandrogenic drugs, or other medications within specified time periods
Pregnancy or not using an adequate form of birth control
Involvement in the planning and/or conduct of the study (applied to any Quintiles or AstraZeneca employee and their close relatives and/or
staff at the study site, regardless of their role in accordance with their internal procedures)
Previous randomization to treatment in the present study
Inability to understand or cooperate with the requirements of the study
Was legally or mentally incapacitated
29
Reporting suicidal ideation of Type 4 or 5 in the past 2 months or suicidal behaviour in the past 6 months as measured by the C-SSRS at
baseline
ECG, electrocardiogram; FSH, follicle-stimulating hormone; LH, luteinizing hormone; LLN, lower limit of normal; ULN, upper limit of normal.
30
533
534
Supplemental Table 2. Baseline-adjusted changes in LH AUC, total testosterone Cavg,, free testosterone Cavg, LH AUC:FSH AUC ratio, and LH pulsatility
parameters relative to placebo
Parameter Day AZD4901 20 mg/day AZD4901 40 mg/day AZD4901 80 mg/day
n n n
Luteinizing hormone AUC
All analyzed patients7 13 −12.96 (−41.48 to 29.47) 14 −21.24 (−46.59 to 16.16) 15 −52.01 (−67.27 to −29.64)
28 12 −23.96 (−50.12 to 5.94) 12 −3.09 (−36.39 to 47.64) 14 −24.31 (−49.56 to 13.58)
Non-ovulating patientsa7 10 −2.38 (−31.94 to 40.01) 12 −15.94 (−40.44 to 18.64) 13 −46.44 (−61.80 to −24.90)
28 9 −13.87 (−41.49 to 26.79) 11 −2.12 (−32.23 to 41.36) 12 −34.87 (−54.57 to −6.62)
Total testosterone Cavg
All analyzed patients7 14 −7.53 (−23.32 to 11.51) 16 6.64 (−22.13 to 11.93) 15 −28.65 (−40.88 to −13.89)
28 14 −0.09 (−17.35 to 20.77) 13 −3.57 (−20.54 to 17.01) 14 −17.01 (−31.55 to 0.63)
Non-ovulating patientsa7 11 −4.45 (−20.03 to 14.16) 14 −1.93 (−17.18 to 16.12) 13 −27.10 (−38.73 to −13.26)
28 11 0.09 (−16.45 to 19.90) 12 −4.05 (−19.69 to 14.64) 12 −20.83 (−33.80 to −5.32)
31
535
536
Free testosterone Cavg
All analyzed patients7 13 −3.06 (−21.77 to 20.13) 15 2.42 (−16.65 to 25.84) 14 −19.20 (−34.62 to −0.14)
28 14 6.34 (−14.12 to 31.68) 12 10.72 (−11.30 to 38.21) 13 −16.68 (−33.05 to 3.68)
Non-ovulating patientsa7 10 −2.72 (−20.32 to 18.75) 14 −1.87 (−18.11 to 17.58) 12 −22.77 (−36.00 to −6.80)
28 11 11.62 (−8.46 to 36.10) 12 4.87 (−13.42 to 27.03) 11 −23.77 (−37.31 to −7.32)
LH number of pulses/8 hoursa
All analyzed patients7 13 −1.33 (−2.89 to 0.23) 12 −1.18 (−2.83 to 0.46) 14 −3.55 (−5.10 to −2.00)
28 12 −0.42 (−2.08 to 1.24) 11 −0.90 (−2.62 to 0.81) 13 −1.19 (−2.85 to 0.46)
Non-ovulating patientsa7 10 −1.58 (−3.13 to −0.03) 10 −1.40 (−2.97 to 0.18) 12 −3.90 (−5.38 to −2.42)
28 9 −0.14 (−1.80 to 1.51) 10 −1.18 (−2.81 to 0.45) 11 −1.89 (−3.48 to −0.30)
LH basal secretion
All analyzed patients7 13 −19.84 (−63.74 to 77.22) 12 −32.30 (−70.10 to 53.26) 14 −78.83 (−90.34 to −53.61)
28 12 −2.80 (−58.29 to 126.55) 11 −3.52 (−59.40 to 129.24) 13 −37.16 (−72.85 to 45.44)
Non-ovulating patientsa7 10 −26.57 (−65.94 to 58.31) 10 −20.35 (−63.13 to 72.06) 12 −80.56 (−90.65 to −59.57)
28 9 −0.04 (−56.36 to 128.94) 10 −16.68 (−62.84 to 86.80) 11 −61.32 (−82.41 to −14.97)
32
LH mass-per-pulse
All analyzed patients7 13 58.95 (−5.88 to 168.41) 12 41.33 (−17.05to 140.80) 14 14.88 (−31.39 to 92.37)
28 12 0.47 (−42.81 to 76.50) 11 11.48 (−37.17 to 97.78) 13 −4.24 (−44.85 to 66.26)
Non-ovulating patientsa7 10 65.57 (−2.06 to 179.90) 10 69.20 (−0.08 to 186.51) 12 58.56 (−3.99 to 161.88)
28 9 −3.18 (−45.08 to 70.68) 10 16.52 (−33.02 to 102.70) 11 −10.25 (−47.68 to 53.96)
LH AUC:FSH AUC ratio†
All analyzed patients7 11 −0.00 (−0.50 to 0.49) 14 −0.31 (−0.78 to 0.15) 13 −0.70 (−1.17 to −0.23)
28 11 −0.44 (−0.96 to 0.07) 11 −0.47 (−0.98 to 0.04) 13 −0.72 (−1.21 to −0.23)
Non-ovulating patientsa7 10 −0.08 (−0.59 to 0.44) 12 −0.28 (−0.77 to 0.20) 11 −0.66 (−1.16 to −0.17)
28 9 −0.30 (−0.86 to 0.26) 10 −0.44 (−0.98 to 0.09) 11 −0.87 (−1.39 to −0.35)Data are baseline-adjusted percentage changes in geometric means relative to placebo (95% confidence interval) unless otherwise stated. Values in bold
correspond to nominal P < .05.
aPatients with no biochemical evidence of ovulation (serum progesterone <6 ng/dL [19.1 nmol/L] at all study visits).
bBaseline-adjusted differences in arithmetic means relative to placebo (95% confidence interval).
AUC, area under the concentration–time curve (0–8 hours post-dose); C avg, average concentration over 8 hours post-dose; FSH, follicle-stimulating hormone;
LH, luteinizing hormone.
33
537
538
539
540
541
542
Supplemental Table 3. Pharmacokinetic parameters for AZD4901 and is metabolite AZ12592232
Parameter AZD4901 20mg/day AZD4901 40mg/day AZD4901 80mg/day
Day 7 Day 28 Day 7 Day 28 Day 7 Day 28
AZD4901
AUC (hng/mL) 1590 (585) 1690 (517) 2440 (802) 2500 (741) 4400 (1400) 3800 (794)
Cmax (ng/mL) 335 (128) 342 (104) 497 (148) 471 (110) 835 (213) 732 (155)
Tmax (h) 2.11 (0.85) 1.82 (0.72) 1.54 (0.57) 1.81 (0.63) 1.68 (0.37) 1.90 (1.09)
AZ12592232
AUC (hng/mL) 607 (275) 605 (191) 1150 (485) 1090 (445) 2120 (709) 1880 (446)
Cmax (ng/mL) 88.4 (35.2) 87.4 (25.0) 166 (66.4) 157 (58.9) 315 (130) 276 (76.5)
Tmax (h) 5.57 (1.74) 5.86 (1.99) 2.18 (2.02) 3.65 (2.35) 2.77 (2.09) 2.43 (2.08)
AZ12592232:AZD4901
ratio
AUC 0.38 (0.098) 0.37 (0.078) 0.46 (0.134) 0.44 (0.136) 0.48 (0.118) 0.50 (0.111)
Cmax 0.27 (0.068) 0.26 (0.051) 0.33 (0.092) 0.33 (0.082) 0.37 (0.089) 0.38 (0.073)
34
543
Data are arithmetic mean (standard deviation).AUC, area under the concentration–time curve (0–8 hours post-dose); C max, maximum concentration; Tmax, time
to maximum concentration.
35
544
545
Supplemental Table 4. Summary of biochemical measurements at baseline and on days 7 and 28
AZD4901 20mg/day AZD4901 40 mg/day AZD4901 80mg/day Placebo
Follicle stimulating hormone Cavg (IU/L)
Baseline 6.15 ± 2.06 (13) 4.57 ± 1.67 (15) 4.52 ± 1.49 (14) 4.68± 1.35 (13)
Day 7 6.48 ± 1.98 (13) 4.73 ± 1.07 (15) 3.79 ± 1.45 (14) 5.12 ± 1.35 (16)
Day 28 5.30 ± 1.53 (13) 5.28 ± 1.21 (12) 5.25 ± 1.44 (13) 4.53 ± 1.49 (15)
Luteinizing hormone AUC (hourIU/L)
Baseline 78.3 ± 28.0 (13) 73.0 ± 36.7 (15) 73.7 ± 30.1 (15) 72.7 ± 40.6 (13)
Day 7 91.9 ± 79.8 (14) 66.2 ± 30.4 (15) 47.9 ± 36.9 (15) 78.8 ± 40.0 (16)
Day 28 81.0 ± 63.7 (13) 74.1 ± 31.9 (13) 55.4 ± 29.2 (14) 73.2 ± 40.3 (14)
Luteinizing hormone mass per pulse (IU/L)
Baseline 9.19 ± 3.98 (13) 10.0 ± 6.58 (14) 9.93 ± 5.46 (14) 7.86 ± 5.57 (13)
Day 7 15.0 ± 11.3 (14) 10.4 ± 5.37 (13) 12.0 ± 8.53 (15) 8.85 ± 6.45 (16)
Day 28 13.2 ± 9.61 (13) 11.1 ± 6.97 (13) 10.2 ± 6.16 (13) 9.18 ± 5.41 (14)
36
546
Luteinizing hormone pulse frequency (number of pulses/8 hours)
Baseline 6.38 ± 2.14 (13) 5.50 ± 2.35 (14) 5.79 ± 2.08 (14) 7.15 ± 2.27 (13)
Day 7 5.50 ± 1.99 (14) 5.85 ± 1.68 (13) 3.73 ± 2.09 (15) 6.75 ± 2.57 (16)
Day 28 6.00 ± 2.42 (13) 5.38 ± 1.33 (13) 5.15 ± 1.63 (13) 6.21 ± 2.39 (14)
Luteinizing hormone basal secretion (IU/L)
Baseline 147 ± 57.0 (13) 146 ± 107 (14) 123 ± 84.0 (14) 155 ± 94.6 (13)
Day 7 145 ± 151 (14) 105 ± 74.4 (13) 63.4 ± 96.8 (15) 153 ± 85.9 (16)
Day 28 153 ± 134 (13) 125 ± 85.3 (13) 64.8 ± 52.1 (13) 123 ± 102 (14)
Total testosterone Cavg (nmol/L)
Baseline 1.96 ± 0.624 (14) 2.07 ± 0.921 (16) 2.25 ± 0.616 (15) 1.68 ± 0.680 (16)
Day 7 1.82 ± 0.544 (14) 2.03 ± 0.994 (16) 1.65 ± 0.582 (15) 1.77 ± 0.763 (16)
Day 28 1.97 ± 0.771 (14) 2.02 ± 0.761 (13) 1.81 ± 0.601 (14) 1.72 ± 0.807 (15)
37
Free testosterone Cavg (pmol/L)
Baseline 66.3 ± 37.8 (14) 72.3 ± 32.7 (15) 91.9 ± 30.1 (14) 84.5 ± 57.4 (16)
Day 7 60.8 ± 33.2 (13) 70.2 ± 32.5 (16) 70.2 ± 32.2 (14) 82.5 ± 64.5 (16)
Day 28 70.4 ± 40.4 (14) 76.0 ± 30.2 (13) 74.9 ± 32.5 (14) 83.7 ± 61.8 (15)
Estradiol concentration (pmol/L)
Baseline 240 ± 135 (15) 300 ± 158 (17) 322 ± 242 (17) 254 ± 169 (16)
Day 7 245 ± 154 (15) 226 ± 180 (17) 252 ± 199 (17) 190 ± 55.5 (16)
Day 28 246 ± 137 (15) 202 ± 188 (14) 248 ± 266 (16) 220 ± 100 (16)
Progesterone concentration (nmo/L)
Baseline 2.75 ± 2.63 (14) 4.23 ± 5.89 (17) 9.74 ± 14.7 (16) 3.16 ± 5.33 (16)
Day 7 4.16 ± 7.69 (15) 3.42 ± 5.36 (16) 4.77 ± 6.85 (17) 1.47 ± 0.927 (16)
Day 28 4.03 ± 8.05 (15) 1.66 ± 0.97 (13) 1.48 ± 1.59 (15) 6.90 ± 12.3 (16)
38
Thyroid stimulating hormone concentration (mU/L)
Baseline 2.62 ± 1.03 (15) 2.73 ± 1.91 (17) 2.24 ± 0.826 (15) 2.06 ± 0.961 (16)
Day 7 2.28 ± 0.733 (15) 2.56 ± 1.30 (16) 2.49 ± 1.05 (17) 1.58 ± 0.579 (16)
Day 28 2.40 ± 0.948 (15) 2.47 ± 1.24 (12) 2. 74 ± 1.57 (15) 1.65 ± 0.584 (16)
Prolactin concentration (mIU/L)
Baseline 239 ± 75.9 (14) 291 ± 169 (17) 192 ± 70.9 (17) 221 ± 100 (16)
Day 7 234 ± 110 (15) 240 ± 142 (16) 213 ± 98.1 (17) 179 ± 60.0 (16)
Day 28 211 ± 61.2 (15) 232 ± 84.0 (14) 200 ± 61.6 (15) 197 ± 79.6 (16)
Glycated hemoglobin (%)
Baseline 4.93 ± 0.327 (15) 5.25 ± 0.416 (17) 5.24 ± 0.255 (17) 5.14 ± 0.361 (16)
Day 7 4.87 ± 0.32 (14) 5.27 ± 0.447 (16) 5.19 ± 0.232 (16) 5.10 ± 0.410 (14)
Day 28 4.81 ± 0.316 (14) 5.26 ± 0.503 (14) 5.04 ± 0.253 (16) 4.95 ± 0.372 (15)
39
Glycated hemoglobin (mmol/mol)a
Baseline 30 ± 3.6 (15) 34 ± 4.6 (17) 33 ± 2.8 (17) 32 ± 3.9 (16)
Day 7 29 ± 2.2 (14) 34 ± 4.9 (16) 33 ± 2.5 (16) 32 ± 4.5 (14)
Day 28 29 ± 3.5 (14) 34 ± 5.5 (14) 31 ± 2.7 (16) 31 ± 4.0 (15)
Data are arithmetic mean±standard deviation (number of patients for whom data were available).
aValues calculated from arithmetic means and standard deviations of glycated hemoglobin (%) in Supplemental Table 4 using the formula HbA 1c(mmol/mol)
= 10.93HbA1c(%) – 23.5 (Hoelzel W, Weykamp C, Jeppsson JO, Miedema K, Barr JR, Goodall I, Hoshino T, John WG, Kobold U, Little R,
Mosca A, Mauri P, Paroni R, Susanto F, Takei I, Thienpont L, Umemoto M, Wiedmeyer HM. IFCC reference system for measurement of
hemoglobin A1c in human blood and the national standardization schemes in the United States, Japan, and Sweden: a method-comparison study.
Clin Chem. 2004; 50:166–174).
AUC, area under the concentration–time curve (0–8 hours post-dose); Cavg, average concentration (0–8 hours post-dose).
40
547
548
549
550
551
552
553
Supplemental Table 5. Number of patients dosed with AZD4901 who reported adverse events, by preferred term
Adverse event AZD4901
20 mg/day
AZD4901
40 mg/day
AZD4901
80 mg/day
Placebo Total
(n = 15) (n = 17) (n = 17) (n = 16) (n = 65)
Headache 2 3 4 5 14
Nasopharyngitis 2 1 1 1 5
Dizziness 1 1 1 3
Nausea 1 1 2
Vaginal hemorrhage 1 1 2
Muscle spasms 1 1 2
Increased hepatic enzyme 2 2
Rash 2 2
Acne 1 1 2
Influenza-like illness 1 1 2
Upper respiratory tract infection 1 1 2
41
554
Abdominal pain 1 1 2
Upper abdominal pain 1 1 2
Constipation 1 1
Vulvovaginal mycotic infection 1 1
Pelvic pain 1 1
Muscle strain 1 1
Conjunctivitis 1 1
Lower abdominal pain 1 1
Sleep apnea syndrome 1 1
Procedural dizziness 1 1
Tooth fracture 1 1
Allergic rhinitis 1 1
Ear pain 1 1
Appendicitis 1 1
Nodule 1 1
42
Migraine 1 1
Presyncope 1 1
Syncope 1 1
Abdominal distension 1 1
Diarrhea 1 1
Increased alanine aminotransferase level 1 1
Increased aspartate aminotransferase level 1 1
Nephrolithiasis 1 1
Pyrexia 1 1
Hot flush 1 1
Vaginal discharge 1 1
Gastrointestinal infection 1 1
A patient may have reported the same adverse event several times.
43
555
Supplemental Table 6. Health-related quality of life assessed using the 36-item Short-Form Health Survey
Domain AZD4901 20mg/day AZD4901 40mg/day AZD4901 80mg/day Placebo
Baseline Day 28 Baseline Day 28 Baseline Day 28 Baseline Day 28
Physical Functioning 51.3 (9.7) 54.3 (4.0) 47.8 (11.3) 51.2 (10.7) 55.7 (3.4) 55.6 (3.8) 54.5 (3.5) 55.6 (2.3)
Role Physical 52.8 (7.1) 54.4 (4.1) 54.8 (4.2) 53.6 (5.1) 54.1 (6.6) 55.5 (4.4) 53.6 (9.7) 55.3 (2.5)
Bodily Pain 54.4 (10.7) 56.5 (7.5) 55.5 (8.1) 53.9 (8.1) 58.3 (6.2) 57.3 (5.6) 58.1 (6.4) 57.0 (4.5)
General Health 50.4 (11.3) 51.4 (10.5) 50.2 (8.0) 51.0 (6.9) 53.3 (8.0) 55.5 (8.5) 53.6 (6.1) 52.4 (4.6)
Vitality 53.1 (10.9) 51.9 (10.4) 53.5 (8.4) 53.3 (9.1) 53.4 (8.1) 56.3 (6.9) 52.1 (8.7) 53.7 (7.3)
Social Functioning 53.6 (7.4) 55.3 (4.5) 53.5 (5.2) 53.5 (7.9) 54.1 (6.3) 54.3 (6.1) 51.4 (10.0) 53.8 (6.3)
Role Emotional 53.8 (6.0) 55.3 (2.2) 53.2 (4.3) 53.2 (7.5) 55.0 (1.7) 55.9 (0.0) 47.9 (14.6) 53.9 (4.3)
Mental Health 54.1 (6.5) 53.8 (6.8) 56.1 (3.0) 54.8 (9.9) 55.4 (6.7) 55.6 (5.1) 54.2 (8.4) 53.4 (6.8)
Data are arithmetic mean (standard deviation). Each score is calculated on a scale of 0 to 100, with lower scores indicating poorer quality of life.
Improvement in HRQoL is reflected in an increase in score from baseline.
HRQoL, health-related quality of life.
44
556
557
558
559