The NK3 Receptor Antagonist ESN364 Interrupts Pulsatile LH Secretion and Moderates Levels of Ovarian Hormones Throughout the Menstrual Cycle

Women’s health disorders such as uterine fibroids and endometriosis are currently treated by GnRH modulators that effectively suppress the hypothalamic-pituitary-gonadal axis. The neurokinin-3 receptor (NK3R) is an alternative target with an important role in the modulation of this axis. In this report, we demonstrate that systemic administration of an NK3R antagonist (ESN364) prolongs the LH interpulse interval in ovarectomized ewes and significantly lowers plasma LH and FSH concentrations in castrated nonhuman primates (Macaca fascicularis). Moreover, daily oral dosing of ESN364 throughout the menstrual cycle in M fascicularis lowered plasma estradiol levels in a dose-dependent manner, although nadir levels of estradiol were maintained well above menopausal levels. Nevertheless, estradiol levels during the follicular phase were sufficiently inhibited at all doses to preclude the triggering of ovulation as evidenced by the absence of the LH surge and failure of a subsequent luteal phase rise in plasma progesterone concentrations, consistent with the absence of normal cycle changes in the uterus. Apart from the point at surge, FSH levels were not altered over the course of the menstrual cycle. These effects of ESN364 were reversible upon cessation of drug treatment. Together these data support the proposed role of neurokinin B-NK3R signaling in the control of pulsatile GnRH secretion. Furthermore, in contrast to GnRH antagonists, NK3R antagonists induce a partial suppression of estradiol and thereby offer a viable therapeutic approach to the treatment of ovarian sex hormone disorders with a mitigated risk of menopausal-like adverse events in response to long-term drug exposure.

Modulation of the hypothalamic-pituitary-gonadal (HPG) axis is a clinically validated therapeutic approach for the treatment of disorders of the reproductive system, as proven by the established medical use of GnRH modulators (1). However, the approved duration of use of GnRH modulators is restricted due to their castrating effects and consequent menopausal-like adverse events, including loss of bone mineral density and incidences of hot flashes (2). Therefore, alternative approaches that offer a more refined modulation of the HPG axis would be of therapeutic benefit.

Genetic studies in human populations have revealed that a loss-of-function mutation in the gene encoding the neurokinin 3 receptor (NK3R; TACR3), or the gene encoding its cognate ligand neurokinin B (NKB) (TAC3), displays a phenotype of hypogonadotropic hypogonadism (3, 4). Unlike the consequences of analogous mutations in the GnRH or kisspeptin (G protein-coupled receptor 54) receptors (5, 6), however, a loss-of function mutation in TACR3 presents a more subtle phenotype whereby some adult patients are able to recover spontaneous function of the HPG axis including normal menstrual cycling and fertility (7, 8). Furthermore, such patients exhibit differential effects on gonadotropin levels, such that plasma LH levels are profoundly diminished but FSH levels are not significantly different from that of normal individuals (9), suggesting that this phenotype arises from a lowered GnRH pulse frequency (10). The target validation imparted by these genetic studies and the presentation of this refined phenotype inspired us to examine the pharmacology of NK3R antagonists with application toward the treatment of reproductive health disorders in women. This research culminated in the invention of the clinical development compound, ESN364 (11, Figure 1).

Figure 1

Chemical structure of ESN364, (R)-(4-fluorophenyl)(8-methyl-3-(3-methyl-1,2,4-thiadiazol-5-yl)-5,6-dihydro-[1,2,4]triazolo[4,3-a]pyrazin-7(8H)-yl)methanone.

NK3R antagonists have been assessed previously in the search for a treatment of schizophrenia. All clinical development candidates were discontinued due to lack of efficacy, which was interpreted as a consequence of poor drug availability at the central sites of action (12). Our comparison of the effects of these previous generation NK3R antagonists with ESN364 in the modulation of the HPG axis has demonstrated that compound efficacy in lowering plasma LH in the castrate rat correlates well with the potency-normalized free drug concentrations in plasma and brain and that this aspect is key to the superior profile of ESN364 in modulating the HPG axis (11, 13).

Various studies have implicated NK3R signaling in the arcuate nucleus in the modulation of pulsatile GnRH secretion (14–17). However, these studies were restricted to acute, direct infusion of test compounds into selected brain regions due to the limited pharmacokinetic profile of the agonists and antagonists used. In contrast, the oral efficacy of ESN364 permits the investigation of the pharmacology of an NK3R antagonist upon systemic exposure and, furthermore, allows for the maintenance of dosing over extended periods. Here we report that the systemic administration of an NK3R antagonist prolongs the LH interpulse interval and thereby lowers plasma LH but without effect on corresponding baseline FSH levels in intact females, analogous to the phenotype in patients with mutations inactivating TAC3 or TACR3. Furthermore, we demonstrate drug effects over the course of the menstrual cycle in cynomolgus monkeys (Macaca fascicularis), in which female sex hormone levels are lowered but the levels of estradiol remain nonmenopausal. These data, together with the preclinical drug profile of ESN364 presented previously (11), comprehensively support the concept of NK3R antagonists as a novel approach for the treatment of women’s health disorders.

Materials and Methods

Evaluation of gonadotropins and thermoregulation in the ovariectomized (OVX) ewe

Ten Corriedale ewes (body weight 56.6 ± 3.4 kg), 3–4 years of age were ovariectomized according to standard operating procedures as previously described (18) and approved by Monash University Ethics Committee. After 4 months recovery, animals were acclimatized to housing in single pens for a period of 7 days with ad libitum access to water and chaffed lucerne hay. One day prior to experimentation, the animals received a jugular vein cannula (Dwellcath; Tuta Laboratories). The cannulae were kept patent with heparinized saline and blood sampling commenced at 7:00 am. Jugular venous blood samples (5 mL) were drawn every 10 minutes for 8 hours, placed into heparinized plastic tubes, and centrifuged at 2500 × g for 15 minutes. The plasma was decanted off the samples into parallel vials designated for the LH/FSH measurement or pharmacokinetic analysis, respectively. Plasma samples were frozen and stored at −15°C until assay. Rectal temperatures were also monitored with a probe at hourly intervals throughout the experiment, starting at 7:40 am.

On the day of the experiment, the NK3R antagonist ESN364 was formulated in physiological saline with 9% 2-hydroxypropyl-β-cyclodextrin at a concentration of 2 mg/mL. At 240 minutes after the initiation of blood sampling, ESN364 (1 mg/kg, n = 5) or vehicle (n = 5) was administered by an iv bolus injection at a dose volume of 0.5 mL/kg through the jugular cannulae, and the injected material was flushed into the animal with 5 mL of heparinized saline. Blood sampling resumed at the indicated intervals following this iv administration.

Duplicate aliquots (100 μL) of plasma samples from individual animals were assayed for LH and FSH, respectively. In the case of LH, radioimmunoassay (RIA) and analysis were performed as previously described (19). Two assays were performed with high repeatability using the assay standard of NIH-oLH-S18 with a sensitivity of 0.1 ng/mL, and the intra- and interassay coefficients of variation (CVs) were 5.5% and 17%, respectively. A pulse analysis was performed according to published procedures (20). Mean plasma levels, interpulse interval, and pulse amplitude were quantified for each animal across three periods. The average data for each ewe were used to obtain means (±SEM) for each group (ESN364 treated and vehicle treated), which were compared using a repeated-measures ANOVA. For FSH, the assay was performed as previously described (21) but in this case using NIH-oFSH-SIAFP-RP-2 as the assay standard with a sensitivity of 0.6 ng/mL and intra- and interassay CVs of 4.7% and 10%, respectively. Two assays were performed with confirmed repeatability of the measurements. Plasma concentrations of ESN364 were determined by liquid chromatography and tandem mass spectrometry (LCMS/MS) with a lower limit of quantification of 26 nM, and data were analyzed by noncompartmental methods using PK Solutions 2.0 software (Summit Research Services).

Evaluation of gonadotropins after subchronic, oral dosing of ESN364 in castrated monkey

This study was conducted in accordance with an International Animal Care and Use Committee-approved protocol according to the guidelines for the humane and ethical use of laboratory animals. The maintenance and handling of the monkeys used in this study was outsourced to PHARMARON. Sexually mature, male, cynomolgus monkeys (M fascicularis; n = 4, aged 4.5–6.1 y, body weight 4.5 ± 0.5 kg) were castrated and allowed to recover for more than 6 months prior to this experiment. Monkeys were group housed and maintained on an alternating 12-hour light, 12-hour dark cycle on a standard laboratory chow diet supplemented with fruit. Water was provided ad libitum. A pilot study was performed during which monkeys were treated with vehicle (0.5% methylcellulose/water), and blood samples were collected at the time intervals specified below. Animals were given a 1-week recovery period before entry into the main study. ESN364 was formulated in 0.5% methylcellulose/water and administered by oral gavage at a dose of 5 mg/kg (dose volume 5 mL/kg). Daily dosing occurred at 8:00 am for a period of 5 consecutive days. Blood samples were collected by venepuncture on the first and final day of dosing at −1, −0.5, 0 (predose), 0.5, 1, 1.5, 2.5, 5, 8, 12, and 24 hours.

Samples were collected into centrifuge tubes containing K₃EDTA and centrifuged at 2500 × g for 15 minutes. The plasma was decanted off the samples into parallel vials designated for LH, FSH, or pharmacokinetic analysis, respectively. All sample vials were immediately frozen and stored at −20°C until the assay. Plasma LH and FSH were measured using homologous (macaque) RIAs as described previously (22, 23). Recombinant cynomolgus LH (AFP6936A) and recombinant cynomolgus FSH (AFP6940A) were used as the standard and radiolabeled [¹²⁵I]tracer for the LH and FSH assay, respectively. The sensitivity of the LH and FSH assays ranged between 0.29–1.43 ng/mL and 0.45–1.08 ng/mL, respectively, and the mean intra- and interassay CVs for both LH and FSH were 5% and 12%, respectively. ESN364 plasma concentrations were determined by LCMS/MS with a lower limit of quantification of 3.6 nM, and data were analyzed by noncompartmental methods using WinNonlin Pro (Pharsight Corp).

Evaluation of gonadotropins and ovarian hormones over the menstrual cycle in monkey after subchronic, oral dosing of ESN364

This study was carried out by Covance Laboratories GmbH (Münster, Germany) in compliance with the German Animal Welfare Act and with approval by the local Institutional Animal Care and Use Committee. Cynomolgus monkeys (M fascicularis) were selected as a relevant species for these studies because of the similarity of these nonhuman primates to humans with regard to having a menstrual cycle of similar duration and clear demarcation of the follicular and luteal phases, divided by ovulation (24). Females in the weight range of 3–6 kg and age range of 6–9 years at predose were used. Sexual maturity of females was confirmed prior to the study by at least two menstrual bleedings on 2 consecutive days with 25–43 days in between; animals were randomly divided into groups in which the mean cycle lengths (30–32 d) did not significantly differ between the groups during the observation period. Animals were pair housed in a climate-controlled room (room temperature 19°C–25°C, humidity 40%–70%, eight air changes per hour) with artificial lighting controlled automatically on a 12-hour light, 12-hour dark cycle. Animals were fed twice daily a commercial pellet diet for primates (ssniff P10; ssniff Spezialdiaten GmbH) supplemented regularly with fresh fruit and bread. Water was provided ad libitum. ESN364 was formulated in 0.5% methylcellulose/water and administered by oral gavage (dose volume 5 mL/kg) at daily doses of 10, 25, or 50 mg/kg; a vehicle-treated group was maintained on the same dosing schedule. Daily dosing occurred between 7:00 am and 10:00 am for a period of 35 consecutive days, during which the dosing was initiated on day 2 of the menstrual cycle for individual animals. Each treatment group was comprised of four monkeys.

For the evaluation of hormones, blood samples were collected from all female animals during the dosing phase on days 4, 7, 10, 11, 12, 13, 14, 15, 16, 18, 20, 22, 24, and 27 to provide ample coverage of the menstrual cycle. Approximately 2.5 mL of blood was collected from the vena cephalica antebrachii between 7:00 am and 10:00 am (predosing) at each interval to yield 1 mL of serum. Serum samples were split into two portions and kept frozen at −20°C until analysis. All hormone assays were validated in-house according to good laboratory practice regulations by Covance Laboratories GmbH for use in cynomolgus monkey, and the data were within historical ranges for this laboratory, as previously reported (24). RIAs were performed using commercially available assay kits for estradiol (DSL-4800assay kit; Beckman Coulter; 74889 Sinsheim; assay sensitivity of 20.3 pmol/L, intra- and interassay CVs of 11.3% and 18%, respectively) and progesterone (TKPG assay kit; Siemens GmbH; 65760 Eschborn; assay sensitivity of 0.06 nmol/L, intra- and interassay CVs of 6.8% and 9.1%, respectively). The RIA for FSH was performed by competitive binding vs ¹²⁵I-FSH (10 μCi, NEX-173; PerkinElmer Life Science) for the high-affinity, mouse antihuman β-FSH primary monoclonal antibody (Immunotech 0373; Beckman Coulter Immunotech) and precipitation of the antigen-antibody complex was accomplished using the secondary antibodies polyclonal rabbit antimouse Ig (DAKO; Z-0259) and antimouse Ig (ICN 55861; MP Biomedicals). Bound and unbound ¹²⁵I-FSH were separated by centrifugation in which the amount of radioactivity recovered was inversely related to the amount of FSH present in the sample in reference to FSH standards of known concentration assayed in parallel. The FSH assay had a sensitivity of 0.47 ng/mL and intra- and interassay CVs of 8.60% and 8.45%, respectively. LH was measured by mouse Leydig cell assay (25), a LH bioassay wherein serum samples are incubated with freshly prepared mouse Leydig cells and consequent T production is measured by RIA and interpreted relative to a standard curve for the response to LH. The LH bioassay sensitivity was 0.9 IU/L with an established working range from 0.176 to 4500 IU/L and intra- and interassay CVs of 14.3% and 20.6%, respectively.

Blood samples (∼0.8 mL of blood to yield ∼0.3 mL of plasma) for pharmacokinetic analysis were collected from the vena cephalica antebrachii or vena saphena into K₂EDTA tubes on day 1 and day 25 of dosing, only. The interval of blood collection was 0 (predose), 1, 2, 4, 6, 10, and 24 hours after dosing. Blood samples were gently hand mixed and stored on crushed ice prior to centrifugation (1800 × g, 4°C, 10 min) and collected plasma samples were stored frozen at −20°C until analysis. ESN364 plasma concentrations were determined by LCMS/MS (lower limit of quantification of 1 μM), and data were analyzed by noncompartmental methods using Phoenix WinNonlin 6.2 (Pharsight Corp).

Animals were killed on day 36 (1 d after the final dosing, luteal phase of cycle) or in the case of recovery animals on day 64 (28 d after the final dosing). Animals received an im injection with ketamine hydrochloride followed by iv sodium pentobarbitone prior to exsanguination. Uterine tissue from each animal was fixed, embedded in paraffin wax, sectioned at nominal 5 μm, stained with hematoxylin and eosin, and examined microscopically.

Results

Effects of ESN364 on LH pulse dynamics in the OVX ewe

The NK3 antagonist, ESN364 (1 mg/kg, iv bolus), reversibly inhibited the regular, pulsatile secretion of LH in the ovarectomized ewe. The LH interpulse interval, pulse amplitude, and plasma LH levels were unchanged in vehicle-treated animals throughout the duration of the experiment. In comparison, ESN364 treatment repressed the pulse pattern of LH in all treated animals. A comparative example of the LH pulse pattern for a single vehicle-treated vs ESN364-treated ewe is presented in Figure 2, A and B. Overall, ESN364 treatment significantly prolonged the LH interpulse interval to 70 ± 11 minutes (vs 36 ± 6 min, vehicle [VEH]; Figure 2C), significantly reduced plasma LH to 1.4 ± 0.1 ng/mL (vs 2.1 ± 0.4 ng/mL, VEH; Figure 1E), and tended to cause a decrease in LH pulse amplitude to 1.0 ± 0.2 ng/mL (vs 1.6 ± 0.3 ng/mL, VEH; Figure 2D) in the 2-hour period immediately after the dosing. All parameters recovered to baseline levels in the 2- to 4-hour time interval after the dosing. The recovery to baseline LH dynamics corresponded to a decline in ESN364 plasma levels from 5.3 ± 0.9 μM (10 min after the dosing) to 1.5 ± 0.1 μM (2 h after the dosing), as shown in Figure 2H. Plasma FSH levels were unchanged throughout the duration of the experiment in all treatment groups, as shown in Figure 2F.

Figure 2

ESN364 (1 mg/kg, iv bolus) alters LH dynamics and affects core body temperature in the OVX ewe. A and B, LH pulse patterns in individual ewes given an iv bolus injection of vehicle (A) or ESN364 (1 mg/kg, B). C–G, Mean (±SEM) data for vehicle-treated vs ESN364-treated ewes (n = 5/group). Statistical analyses were performed by a two-way ANOVA followed by Sidak’s multiple comparisons test between treatment groups at the indicated time interval. *, P < .05; **, P < .01; ^NSP > .05. H, Corresponding pharmacokinetic profile of ESN364 for this experiment (data ± SEM).

Rectal temperatures were measured at hourly intervals over the duration of the experiment. Four of five ewes in the vehicle-treated group exhibited a transient, relative body temperature increase of 0.7°C or greater at 12:40 pm, in response to the animals being fed at 12:00 pm, consistent with previous reports (26). Thus, at 12:40 pm the body temperature of the vehicle-treated group (39.7°C ± 0.3°C) was significantly higher than that of the ESN364-treated group (38.9°C ± 0.2°C), as presented in Figure 2G.

Subchronic ESN364 treatment lowers plasma LH, but not FSH, in the castrated monkey

Oral efficacy in response to ESN364 (5 mg/kg) was demonstrated in castrated male cynomolgus monkeys (n = 4). In pilot studies in which monkeys were treated only with vehicle (0.5% methylcellulose), plasma LH levels were stable in a range between 9.8 ± 1.4 and 11.9 ± 1.6 ng/mL, and FSH levels were stable in a range between 22.0 ± 1.2 and 23.8 ± 1.9 ng/mL between sampling times at predose (time = 0) and 8 hours after treatment.

After a single treatment with ESN364 (day 1), plasma LH levels declined significantly from baseline (time = 0) levels of 13.2 ± 3.1 ng/mL to trough (time = 2.5 h) levels of 7.1 ± 1.1 ng/mL (Figure 3A). Plasma LH levels recovered fully by 8 hours after treatment (14.0 ± 1.8 ng/mL). The daily response to ESN364 on plasma LH was similar after repeated dosing. Thus, on day 5, baseline (time = 0), trough (time = 2.5 h), and recovery (time = 8 h) levels of plasma LH were 11.2 ± 3.0, 6.3 ± 1.2, and 11.3 ± 2.5 ng/mL, respectively.

Figure 3

Effect of repeated, morning (8:00–9:00 am) dosing of ESN364 (5 mg/kg, oral, QD) on plasma LH (A) and FSH (B) in castrated male, cynomolgus monkey (n = 4). The effect of vehicle treatment on LH and FSH levels was measured 1 week prior to the repeat-dosing experiment wherein gonadotropins were assayed on days 1 and 5. Statistical analyses performed by a two-way ANOVA determined significant differences between vehicle and ESN364 treatment groups for both LH (P < .03) and FSH (P < .01), although significant changes at specific time points were not found in the post hoc multiple comparison test (Dunnett’s). There were no significant differences (after the dosing) between day 1 and day 5 of treatment with ESN364. C, Corresponding pharmacokinetic profile of ESN364 for this experiment. Note that the T = 0 data for day 5 represents predose trough levels (ie, 24 h after the dose from d 4). D, Hysteresis loop for LH on day 1 of this experiment (ie, the change in the PK-PD effect relationship over time, as indicated by the direction of the arrows). Mean data ± SEM are presented in all cases.

In contrast to the modulation of LH, there was no significant daily change in FSH with repeated dosing (Figure 3B). On day 1, ESN364 treatment lowered plasma FSH from baseline (time = 0, 20.6 ± 2.1 ng/mL) to nadir (time = 2.5 h, 14.3 ± 0.7 ng/mL); however, these nadir levels of FSH were maintained for 24 hours up to the next daily dosing. For example, FSH levels at predose on day 5 (time = 0, 17.4 ± 1.5 ng/mL) were lower than that for vehicle-treated baseline (time = 0, 23.8 ± 1.9 ng/mL) and the aforementioned levels at predose on day 1. Notably, dosing on day 5 elicited no discernible change in plasma FSH levels over the experimental day with values ranging from 17.2 ± 1.5 to 21.7 ± 3.8 ng/mL.

The corresponding pharmacokinetic data for this experiment revealed similar area under the curve for days 1 and 5, indicative of no significant drug accumulation or alternate changes in pharmacokinetics with repeated dosing of ESN364 (Figure 3C). The mean elimination half-lives (T_½) were determined to be 4.8 ± 1.1 hours and 5.1 ± 1.0 hours for days 1 and 5, respectively. For LH, the daily recovery to baseline levels corresponded to a decline in ESN364 plasma levels below 3.5 μM on both days 1 and 5 of dosing. The hysteresis plot reveals an indirect relationship over time between plasma drug levels (x-axis) and efficacy on plasma LH levels (y-axis; Figure 3D). Thus, although drug-plasma levels plateau at approximately 6 μM after the oral dosing, the percentage of LH inhibition nonetheless continues to increase with time to the extent that maximal LH inhibition (∼45%) occurs when drug-plasma levels are submaximal (∼4.5 μM). This time-dependent discordance between drug-plasma levels and efficacy is consistent with ESN364 having an indirect effect to lower LH levels.

ESN364 treatment blocks the LH surge and decreases ovarian hormone levels throughout the menstrual cycle in monkeys

ESN364 (oral, once daily [QD], 35 d dosing period; dosing initiated on d 2 of cycle) modulated gonadotropin and ovarian hormone levels throughout the menstrual cycle. Basal LH levels (measured 24 h after dosing) were above the lower limit of quantification for all individual monkeys. Basal LH Levels (measured systematically at 24 h after daily dosing) trended lower in ESN364-treated monkeys. For example, in the follicular phase (d −7), the median (±SEM) levels ranged between 8.1 ± 4.6 (50 mg/kg dose group) and 21.0 ± 11.6 IU/L (vehicle), and in the luteal phase (d +8), levels ranged between 9.3 ± 6.6 IU/L (50 mg/kg dose group) and 22.9 ± 10.3 IU/L (vehicle). The data for all graphs are coordinated relative to the day of the LH surge. Because no LH surge was apparent in any of the ESN364-treated monkeys, the data for these monkeys are graphed relative to the mean date of the LH surge observed in vehicle-treated monkeys (eg, treatment d 13, corresponding to cycle d 15). In vehicle-treated monkeys, the LH surge triggering ovulation was observed, as expected, between treatment days 12 and 14 (corresponding to menstrual cycle d 14–16). As mentioned earlier, ESN364 treatment at all dose levels blocked the LH surge (P < .01, vehicle vs all treatment groups; Figure 4A). ESN364 treatment did not inhibit FSH levels throughout the experiment with the exception of the midcycle FSH surge (P < .05, vehicle vs all treatment groups; Figure 4B). Indeed, except for the surge at midcycle, FSH levels in the vehicle-treated group trended lower than that of the ESN364-treated groups over the duration of the experiment, although differences were not statistically significant.

Figure 4

Effect of daily dosing of ESN364 (10, 25, 50 mg/kg, orally) on gonadotropin and ovarian hormone levels throughout the menstrual cycle in sexually mature, cynomolgus monkey (n = 4/group). For each animal, dosing was initiated on day 2 of the menstrual cycle. Mean data ± SEM are presented. A–D, Circulating hormone levels assayed from blood samples collected between 7:00 and 10:00 am on the days indicated; statistical analyses were performed by a one-way ANOVA (on indicated days) followed by Dunnett’s multiple comparisons test for all groups compared with the vehicle group. *, P < .05, **, P < .01. E, Corresponding pharmacokinetic profile for ESN364 in this experiment (data ± SEM); pharmacokinetic sampling was conducted on day 1 (open symbol, broken line) and day 25 (closed symbol, solid line) of the experiment. Note that the T = 0 data for day 25 represents predose trough levels (ie, 24 h after the dose from day 24).

The anticipated rise in estradiol levels was observed in the follicular phase in the vehicle group (Figure 4C). This menstrual pattern for estradiol was diminished by ESN364 treatment (P < .05 on surge d −1 and 0, vehicle vs all treatment groups). There was a dose-dependent trend toward estradiol-lowering effects, most obvious between d −6 to d 0 in the follicular phase, wherein trough levels of estradiol were generally greater than 500 pmol/L for the 10-mg/kg treatment group, as compared with 200–300 pmol/L for the higher-dose treatment groups. Indeed, similar trough levels for estradiol were reached for the 25- and 50-mg/kg treatment groups between days −3 and 9 (relative to LH surge) of the experiment, suggesting that maximal estradiol-lowering effects had been attained over the dose range tested.

Progesterone levels peaked in the luteal phase of the menstrual cycle in the vehicle-treated group, as expected (Figure 4D). In comparison, progesterone levels were significantly attenuated in all ESN364 treatment groups compared with vehicle control (P < .05, for d 3–9 after the surge).

Pharmacokinetic profile of ESN364 in female monkeys

Pharmacokinetic analyses of ESN364 were performed on dosing days 1 and 25 of the same experiment in which gonadotropins and ovarian hormones were tracked (Figure 4E). ESN364 is rapidly absorbed after oral dosing with maximal plasma concentrations achieved within 2 hours of administration (on average) for all dose levels. Exposure (area under the curve_0-t) and maximal plasma concentrations increased essentially linearly with dose. The nearly superimposable curves for days 1 and 25 for all dose groups illustrate that there was no significant accumulation or modulation of pharmacokinetic processes consequent to repeated dosing of ESN364 at the stated dose levels. The corresponding pharmacokinetic half-life values (T_½) were found to be in the range of 4.3–7.4 hours, irrespective of the sampling day. A pharmacokinetic blood sampling was collected 24 hours after dosing (just prior to the next daily dosing) for all treatment groups; these data are not presented on the graph for the 10-mg/kg and 25-mg/kg treatment groups because two of four monkeys yielded ESN364 plasma concentrations below the lower limit of quantification (value of 1 μM) at this time point. However, for the 50-mg/kg treatment group, trough levels of ESN364 were in the range of 8.4–9.3 μM (predose baseline levels and 24 h levels as measured on d 25, respectively). In total, these data confirm that the dose-dependent efficacy on gonadotropins and ovarian hormones may be interpreted commensurate with the expected daily increases in drug exposure.

ESN364 general pharmacology and specific effects on sex organs in female monkeys

All doses were well tolerated and did not evoke any safety concerns. Particular attention was paid to monitoring cardiovascular and respiratory effects (for general safety reasons) as well as neurobehavioral or gastrointestinal (GI) effects as NK3R antagonists have been clinically investigated in the latter two areas (12). Drug treatment did not elicit any changes in electrocardiography assessments, blood pressure, or respiratory rate. Also, there were no significant findings using a standard observation battery for assessment of both peripheral and central nervous system activities. Finally, there were no changes indicative of GI effects based on regular monitoring of food consumption, feces appearance, or signs of nausea.

Necropsy was carried out on three monkeys per treatment group at the conclusion of the 35-day dosing period. The remaining monkey in each treatment group was necropsied after a 28-day recovery period to consider reversibility of any treatment-related findings.

A comprehensive, histopathological analysis of all major tissues revealed the uterus as the primary organ consistently altered by ESN364 treatment. In animals from all ESN364-treated groups, moderate stromal atrophy of the uterus was observed without the presence of normally cycling uterine mucosa. Stromal atrophy was characterized by compact stroma, inactive glands, and no distinction between the zona functionalis and the zona basalis. Representative images collected from a vehicle- and an ESN364-treated monkey (50 mg/kg) are presented in Figure 5, A and B. These effects were reversible as confirmed by normally cycling uterine mucosa in recovery animals from all ESN364-treated groups, indistinguishable from that of control animals. A representative image from an ESN364-treated monkey (50 mg/kg) after recovery is shown in Figure 5C.

Figure 5

Effect of ESN364 on uterus in cynomolgus monkey. Hematoxylin-eosin staining of the uterus in tissue samples collected after 35 days of treatment with vehicle (QD; panel A), ESN364 (50 mg/kg, QD; panel B), or ESN364 (50 mg/kg, QD) followed by a 28-day recovery period (panel C). Note uterine mucosa in the luteal phase in panels A and C with distinction between zona functionalis (F) and zona basalis (B) in contrast to stromal atrophy in panel B.

The only other significant finding at termination was a decrease in ovary weight. Thus, total (unadjusted) ovary weight was 0.563 g in vehicle-treated animals vs 0.293 g (P < .05), 0.398 g (P = N.S.), and 0.267 g (P < .01) in the 10-, 25-, and 50-mg/kg treatment groups, respectively (one way ANOVA, Dunnett’s post hoc analysis). The weight decrease was not associated with any histopathological findings in the ovaries. Neither histopathological nor weight changes were observed for pituitary.

Discussion

The current study presents the pharmacological profile of an orally effective NK3R antagonist, ESN364, on gonadotropin and ovarian hormone secretion under both acute and subchronic dosing paradigms. Previously we have reported that the peripheral administration of NK3R antagonists, including ESN364, decreases plasma LH levels in castrate rat and monkey in a manner demonstrating a clear correlation between pharmacodynamics (PD) and pharmacokinetics (PK) of potency-normalized free drug concentrations in the plasma and brain, respectively (11, 13). Here we extend these findings by demonstrating in OVX ewe that the diminution of plasma LH levels is specifically due to an interruption of the LH pulse. The prolongation of the LH pulse interval observed here in response to iv treatment with an NK3R antagonist is consistent with previous observations made in ewe after the infusion of NK3R antagonists directly into the brain (17, 27). The restoration of the LH pulse in the current study is coincident with the decline of ESN364 plasma levels below approximately 1.5 μM, data that are generally consistent with the PK-PD relationships previously established for ESN364 in monkeys (11).

Ovariectomy in the ewe causes changes in core body temperature considered analogous to menopausal hot flashes (28). In our experiment, the vehicle-treated group expressed a significant elevation in core body temperature in response to feeding that was prevented in the ESN364-treated group, indicative of an antipyrogenic effect. Our findings offer a preclinical corollary to the recent discovery that exogenous administration of NKB induces hot flashes in premenopausal women, the implication being that NK3R antagonists may be a useful remedy for this condition (29). This concept runs counter to the profile of GnRH modulators in which hot flashes are noted as a common adverse event (for example, see reference 30). This difference is perhaps due to GnRH modulators provoking kisspeptin/NKB/dynorphin (KNDy) neuron activity (as a consequence of their inhibition of estrogen feedback control), whereas NK3R antagonists act directly on KNDy neurons to reduce neuronal activity ostensibly by blocking the response to NKB. This interpretation of our data is consistent with the emerging research demonstrating the importance of KNDy neuron activity in the genesis of hot flashes, as illustrated in an OVX rat model in which the elevated core body temperature was either reversed by exogenous administration of estradiol (consistent with its negative feedback on KNDy neurons) or prevented altogether by the ablation of KNDy neurons (31, 32). Thus, NK3R antagonists may be useful in the treatment of hot flashes arising from menopause or from sex hormone-dependent cancer therapies.

Pulsatile LH is used as a surrogate marker of GnRH pulse frequency based on its validation in ewe in which direct measures of both hormones were correlated (33). Thus, the reduced LH pulse frequency shown here is indicative of ESN364 pharmacology to interrupt the GnRH pulse frequency and is consistent with the proposed role of NKB to regulate episodic GnRH secretion through its activation of NK3R expressed on KNDy neurons in the arcuate nucleus (34). Differential GnRH pulse frequencies alter the secretion patterns of the gonadotropins, LH and FSH, with high frequency correlating to greater LH secretion and low frequency correlating to relatively greater FSH secretion (35). The basis of the differential regulation of LH and FSH has been the focus of recent review articles (36, 37).

The experiment in castrate monkeys demonstrated that ESN364 efficacy in lowering plasma LH levels was similar after the acute or repeat dosing, an observation that suggests lack of desensitization to the drug under the repeat-dosing paradigm. Drug efficacy was in step with pharmacokinetics, the latter of which also proved stable over the dosing period. The counterclockwise hysteresis plot (38) is indicative of ESN364 having indirect effects to lower plasma LH levels, a concept consistent with the proposed model that NK3R antagonism at the level of the hypothalamus modulates the GnRH pulse and thereby lowers plasma LH levels. Plasma LH levels recover as ESN364 concentrations decline below approximately 3.5 μM.

In primates, testicular negative feedback via inhibin B is the predominant regulator of FSH secretion over the endogenous hypophysiotropic drive (39, 40). However, in the present experiment, castration is expected to ablate serum inhibin B levels to reveal the component of FSH secretion controlled by pulsatile GnRH release as demarcated by the lower plasma FSH in response to the initial dose of ESN364. Moreover, the low levels of FSH prior to dosing on day 5 may indicate that secretion of FSH was chronically suppressed in the agonadal situation. The basis of this observation is unclear, but it appears unlikely to be entirely due to modulation of GnRH secretion because the corresponding secretion of LH was not affected with similar chronicity. Indeed, repeated daily dosing of ESN364 (as shown on d 5) elicited a daily decrease in plasma LH but not FSH. The latter observation is consistent with the notion that the principal consequence of sustained blockade of NKB-NK3R signaling is a prolongation of the GnRH interpulse interval (41), and it has been shown that the GnRH pulse frequency is a more important determinant of acute LH than FSH secretion in primate (42).

In intact female monkeys, in which the daily dosing was initiated coincident with the menstrual cycle, all dose levels of ESN364 blocked the LH surge. It is unclear whether this response stems from a direct action of ESN364 to antagonize the proposed, obligatory role of NKB signaling in the estrogen-induced surge (43) or is an indirect consequence of drug action to lower basal LH levels and thereby blunt plasma estradiol levels below the threshold concentrations required to trigger the LH surge. There was no statistically significant effect on FSH levels throughout the cycle with the exception of the absence of its midcycle surge. Therefore, in the follicular phase, LH-dependent androgen secretion from theca cells is presumably diminished (in keeping with the previously demonstrated effects of ESN364 to lower plasma LH), providing less substrate to be converted to estrogens by the FSH-dependent aromatase activity in granulosa cells. Furthermore, inhibition of ovulation prevents the formation of the corpus luteum and therefore would consequently abolish this source of estradiol in the luteal phase in women, although this aspect is not demonstrated here because the luteal rise in estradiol is known to be less evident in monkey (24). Overall, this accounts for the dose dependent decline, but incomplete elimination, of estradiol levels throughout the cycle. There was no difference in maximal inhibition of estradiol levels at the 25- or 50-mg/kg dose levels, and in both cases nadir levels of estradiol (≥230 pmol/L) remained well above levels measured in menopausal, cynomolgus monkeys (54 pmol/L) (44). The anticipated rise in progesterone levels in the luteal phase was absent in all ESN364-treated monkeys, consistent with the block of the LH surge and therefore presumed abrogation of the downstream events of ovulation, formation of the corpus luteum, and consequent progesterone secretion. The gonadotropin and sex hormone profiles described here are entirely in step with the histopathology data demonstrating an absence of appearance of normal cycle changes in the uterus of ESN364-treated animals. Histopathological changes in the uterus were reversible when examined 1 month after the cessation of ESN364 treatment. There were no side effects of the drug treatment on cardiovascular, respiratory, GI, or nervous systems over the duration of the study.

The reduction of daily, basal plasma LH but not FSH levels in both the intact female monkey and OVX ewe is consistent with the proposed action of ESN364 to decrease the GnRH pulse frequency. This selective action of NK3R antagonists on the gonadotropins is markedly different from that of the GnRH antagonists wherein the blockade of GnRH receptor signaling diminishes the plasma levels of both LH and FSH (for example, see reference 45).

Additionally, ESN364 pharmacology affirms the role of NKB-NK3R signaling in the regulation of the GnRH pulse generator. There are conflicting reports in the literature as to whether NK3R activation is universally, positively correlated with an increase in GnRH pulse frequency because differential responses to the peptide agonists, eg, NKB and senktide, have provoked discussions on the relevance of sex variability, rodent vs ovine vs primate differences, or changes arising due to the background milieu of sex hormones (46). An alternative hypothesis is that these discrepancies are due to the limited utility of NKB, senktide, and the commercially available NK3R antagonist SB222200, as pharmacological tools owing, at least in part, to the deficient pharmacokinetic and receptor selectivity profiles of these agents. In comparison, our results (references 11 and 13 and additional unpublished observations [G.L.F., H.R.H.]) demonstrate that NK3R antagonists exhibiting adequate levels of potency-normalized free drug brain exposure consistently and selectively diminish LH, indicative of a decrease of the GnRH pulse frequency, in the adult HPG axis in both castrate and intact, male and female, rats and monkeys.

ESN364 has similar oral efficacy and potency to the small-molecule GnRH antagonist Elagolix (NBI-42902) to lower plasma LH in castrated monkeys (47). However, the relevant clinical concern for GnRH modulators is the adverse event profile in which a potential comparative advantage for NK3R vs GnRH antagonists can be envisaged. The blockade of GnRH receptor signaling by a GnRH antagonist nonselectively inhibits both LH and FSH secretion, leading to menopausal-like adverse events with chronic drug exposure. Although the current generation of orally available small molecule GnRH antagonists promises to have an improved safety profile due to the potential to titrate the dose (1), nonetheless, the prospect of menopausal-like side effects clearly remains an ongoing concern (48). By reverse token, the distinct mechanism of action of an NK3R antagonist to interrupt the GnRH pulse frequency and thereby selectively decrease basal LH but not FSH secretion in intact female monkeys is shown here to maintain estradiol above menopausal levels, hence protecting against the occurrence of commonly reported adverse events such as bone mineral density loss and hot flashes.

Furthermore, the precise effect of NK3R antagonists to lower plasma LH without affecting FSH in intact females suggests that NK3R antagonists may be uniquely suited for the treatment of hyperandrogenism, a key aspect of polycystic ovary syndrome (PCOS) (1). PCOS patients express a persistently high LH pulse frequency (indicative of a high GnRH pulse frequency) and consequently a high plasma LH to FSH ratio (49, 50). Under LH regulation, the theca cells produce androgens (mainly androstenedione and T), which diffuse into the granulosa cells and are converted to estrogens by aromatase, an enzyme regulated by FSH. Aromatase activity is saturated in PCOS patients, even though FSH induction of such activity is uncompromised (51), implying that saturation is due to abnormally high levels of androgens in the follicular fluid (52, 53). Thus, the high LH to FSH ratio leads to hyperandrogenism, the latter being a primary diagnostic characteristic of PCOS patients. It follows, therefore, that lowering the LH to FSH ratio with an NK3R antagonist would be expected to decrease levels of androgens more potently than that of estrogens as desired for the treatment of PCOS.

NK3R antagonist pharmacology on ovarian hormones is explained in the context of drug effects in the arcuate nucleus modulating GnRH pulse frequency, a concept consistent with our findings on the LH pulse frequency, the differential effects on LH and FSH in intact females, and the importance of adequate potency-normalized, free drug concentrations in brain and plasma to achieve in vivo efficacy (11, 13). However, it is noted that NKB/NK3R mRNA is also expressed in human mural granulosa cells (54) and the myometrium (55), with significantly elevated expression in leiomyomas (eg, uterine fibroids, [56]). As such, the local effects of NK3R antagonists on ovarian steroidogenesis should also be considered.

In summary, our data indicate that iv and repeated oral administration of the NK3R antagonist ESN364 decreases the LH pulse frequency and lowers plasma LH without significantly affecting FSH levels. The downstream consequence of these actions in females is a lowering of estradiol levels, but not to menopausal levels due to uncompromised FSH-induced estradiol release from granulosa cells. Blockade of NKB-NK3R signaling in the hypothalamus lowers basal LH and estradiol levels and prevents the LH surge, ovulation, and consequent progesterone secretion in the luteal phase of the menstrual cycle. This steroidal sex hormone profile correlates with histopathology findings in the uterus showing an absence of normally cycling uterine mucosa. These data provide strong support for the proposed role of NKB-NK3R signaling in GnRH pulse regulation. Moreover, the moderated estradiol reduction corresponds to a mitigated risk of causing menopausal-like adverse events that advocates the use of NK3R antagonists as an alternative to GnRH ligands in the treatment of reproductive health disorders in women. On this basis, ESN364 has been advanced into clinical trials for the treatment of PCOS and uterine fibroids.

Acknowledgments

We thank our colleagues in the chemistry and pharmacology teams at Euroscreen SA for conducting the studies in support of this work. Also, we thank Stephanie Friderichs-Gromoll (Covance, Münster, Germany) for carrying out the histopathological analysis.

This work was supported by the Ministry of Sustainable Development and Public Works, Walloon Region, Belgium.

Disclosure Summary: G.L.F. and H.R.H. are employed by Euroscreen SA. The other authors have nothing to disclose.

Abbreviations

 
• CV

  coefficient of variation

 
• GI

  gastrointestinal

 
• HPG

  hypothalamic-pituitary-gonadal

 
• KNDy

  kisspeptin/NKB/dynorphin

 
• LCMS/MS

  liquid chromatography and tandem mass spectrometry

 
• NKB

  neurokinin B

 
• NK3R

  neurokinin 3 receptor

 
• OVX

  ovariectomized

 
• PCOS

  polycystic ovary syndrome

 
• PD

  pharmacodynamics

 
• PK

  pharmacokinetics

 
• QD

  once daily.