{"paper_id":"b4ac4c4f-ec8e-4bb2-ad99-ecc6e17ded44","body_text":"Thirteen healthy, premenopausal women, aged 27 to 41 years and with regular menstrual\ncycles (25 to 34 days) were recruited into the study; all provided informed written\nconsent. Mean ± standard error of the mean body mass index was 26.6 ±\n1.6 kg/m 2 . The women were not taking any hormonal contraception nor had an\nintrauterine device  in situ . They had normal physical examination\nand full blood cell count results; renal function, electrolyte levels, liver\nfunction, and electrocardiogram results were within normal limits.\nThe sample size in this and previous studies is based on similar proof-of-concept\nmechanistic studies in our laboratory and by other colleagues, demonstrating\nsignificant changes in reproductive hormone concentrations and in pulsatile LH\nsecretion ( 7 ,  9 ,  23 ). For LH pulse frequency,\npower analysis [based on data reported by Skorupskaite et al ( 20 )] indicates that with  α  = 0.05, eight\nwomen per group would be required to show a significant change with 90% power.\nThe NK3Ra MLE4901 was administered orally at 40 mg twice daily. This dosage of\nMLE4901 reduced LH secretion in healthy women and in women with PCOS ( 20 ,  21 ).\nSchematic presentation of the protocol is shown in  Fig. 1 . All women had a treatment cycle and a no-treatment control\ncycle, with a washout cycle in between if the treatment cycle was first; the order\nof cycles was randomized using sealed envelopes. In the treatment cycle, women\nwere administered the specific NK3Ra MLE4901 orally at 40 mg twice daily for 7\ndays starting on cycle day 5 to 6. Peripheral venous blood was sampled for LH,\nFSH, inhibin B, and estradiol concentration measurement between 08:00 and 10:00\nimmediately before treatment (pretreatment), and on days 2, 4, and 6 of NK3Ra\nadministration immediately before the next dose of NK3Ra was taken\n( i.e. , 12 hours after the previous dose), in the morning after\nthe last dose, and then every 2 to 4 days until ovulation was confirmed by\ntransvaginal ultrasonography (Xario 200, 7.5 MHz probe frequency; Toshiba,\nZoetermeer, The Netherlands). Ovulation was defined either by the last day on\nwhich the preovulatory follicle was seen or the appearance of a corpus luteum. In\nthe control cycle, blood sampling and ultrasound timing were equivalent to that in\nthe treatment cycle. A first-void specimen of urine was collected daily throughout\nthe cycle.\nStudy protocol schematic. The NK3Ra MLE4901 was administered orally to 13\nhealthy women for 7 days starting on cycle day 5 to 6. Reproductive hormones\nwere measured and TV USS was performed pretreatment, on days 2, 4, 6, and 8\nduring the study, and then every 2 to 4 days until ovulation. Urine samples\nwere collected daily until the next menstrual period. LH pulsatility (n = 8)\nwas assessed during 8 hours of blood sampling every 10 minutes on day 6 or 7\nof NK3Ra administration or on the equivalent day of the control cycle.\nReproductive hormones and ultrasound scan findings were compared with those\nof the control cycle, the order of which was randomized. BD, twice daily;\nLMP, last menstrual period; NMP, next menstrual period; TV USS, transvaginal\nultrasonography.\nTransvaginal ultrasonography was used to measure the diameter of the leading\nfollicle, any follicles ≥10 mm in diameter, endometrial thickness, and\nappearance of corpus luteum. Ultrasound scans were performed on the same days as\nassessment of reproductive hormone levels.\nAssessment of LH pulsatility was performed in eight of the 13 women (owing to the\ntime commitment, not all women volunteered for this) on day 6 or 7 of NK3Ra\ntreatment or on cycle day 10 to 12 (the equivalent day) of control cycles. All\nvisits commenced between 08:00 and 09:00. The dose of NK3Ra was administered when\nthe pulsatility assessment was begun; blood samples were collected via an\nindwelling intravenous cannula at 10-minute intervals for 8 hours.\nHematological and biochemical safety monitoring was performed before commencing NK3Ra\ntreatment, at end of drug administration, and 2 to 3 weeks later.\nBlood samples were centrifuged at 4°C for 10 minutes at 3000 rpm and serum was\nfrozen at −20°C or below until analysis. LH and FSH levels were\nmeasured by enzyme-linked immunosorbent assay (ELISA) as previously described ( 23 ) with interassay and intra-assay coefficient\nof variation (CV) <5% at the concentrations measured.\n17 β -estradiol was measured on a Roche Cobas E411\nimmunoassay automated analyzer (Roche Diagnostics, Burgess Hill, United Kingdom). The\nlower limit of quantification was 18.4 pmol/L. The inter- and intra-assay CVs were\n<5% and 6.5%, respectively. Inhibin B was measured by ELISA (Beckman Coulter,\nBrae, CA) with a limit of quantification 2.6 pg/mL and intra-assay CV <8%; all\nsamples were assayed in one run.\nProgesterone level was measured by an inhouse ELISA. The interassay CVs for low and\nhigh progesterone pools, respectively, were 11.4% and 9.1%, and the respective\nintra-assay CVs were 8.9% and 5.6%. The lower limit of detection was 0.1 ng/mL.\nUrinary progesterone concentrations were expressed as a ratio of the creatinine\nconcentration, measured colorimetrically (Alpha Laboratories, Eastleigh, United\nKingdom), and adapted for use on a Cobas Fara centrifugal analyzer (Roche\nDiagnostics, Welwyn Garden City, United Kingdom). Intra-assay CV was <3%; all\nsamples were analyzed in one batch, in duplicate.\nThe number of LH pulses, secretory mass of LH per pulse, and basal (nonpulsatile) and\npulsatile (integral of dual amplitude and frequency regulation) LH secretion were\nidentified by an established deconvolutional algorithm with cluster analysis (93%\nsensitivity and specificity) blinded to treatment allocation, and approximate entropy\nwas quantified as a measure of secretory regularity ( 24 ,  25 ).\nBaseline characteristics between the control and NK3Ra-treated cycles were compared\nby Student paired  t  test (for normally distributed data,\n i.e. , LH, FSH, and estradiol levels; and follicle size) or\nWilcoxon matched-pairs signed-rank test (menstrual cycle day). Serum hormone\nconcentrations and ultrasonography data were compared throughout 7 days of NK3Ra\ntreatment and between the control and treatment groups using repeated measure two-way\nanalysis of variance (ANOVA) followed by Bonferroni multiple comparisons  post\nhoc  analysis. Peak serum hormone concentration, size of the preovulatory\nfollicle, and cycle timing were compared by Student paired  t  test.\nUrinary progesterone concentrations were compared by two-way ANOVA followed by\nBonferroni  post hoc  multiple comparisons test; these data were\navailable for 11 women. Midluteal ( i.e. , LH surge +7 days) urinary\nprogesterone levels and cycle length were compared by Student  t \ntest. Characteristics of LH pulsatile secretion were compared by paired Student\n t  test.\nData are presented as mean ± standard error of the mean. Data not normally\ndistributed were log-transformed before statistical analysis. Differences were\nregarded as significant at a two-sided  P  < 0.05. The\nstatistical software package GraphPad Prism 7 (GraphPad Software, San Diego, CA) was\nused.\nThe study protocol was approved by South East Scotland Research Ethics Committee\n(Reference 09/S1101/67). This study was formally assessed as not being a clinical\ntrial of an investigational medicinal product and, therefore, was not linked to\ntrials databases.\n\nEach woman took part in control and treatment cycles, which were comparable by cycle\nday on which the study had started; baseline serum LH, FSH and estradiol\nconcentrations; the size of the largest ovarian follicle present at that time; and\nthe cycle day for assessment of LH pulsatility ( Table\n1 ).\nBaseline Characteristics of Study Participants (n = 13)\nData given as mean ± SEM.\nAbbreviation: N/A, not applicable.\nControl and NK3Ra-treated cycles were comparable by starting cycle day,\nreproductive hormone levels, and the size of the largest ovarian\nfollicle.\nNK3R antagonism had no effect on LH secretion by single timepoint analysis throughout\n7 days of administration ( Fig. 2A)  or by hourly\nanalysis during 8 hours of frequent LH sampling after dosing on day 6 or 7 of NK3Ra\ntreatment and on the equivalent day of the control cycle ( Fig. 2B) . However, deconvolutional analysis of pulsatile LH\nsecretion, performed in eight of the 13 women on day 6 or 7 of NK3Ra treatment (cycle\nday 10.9 ± 0.2) and on the equivalent day of the control cycle (cycle day 11.1\n± 0.5), showed differences in pulsatile LH secretion between treatment and\ncontrol cycles. An example of an LH pulse frequency profile is shown in  Fig. 3A  and the pulse profile for each participant\nis summarized in  Supplemental Table 1 . LH pulse frequency did not\nchange with NK3Ra treatment [0.69 ± 0.1 pulses/h vs 0.66 ± 0.1 pulses/h\nin control cycles; nonsignificant (ns)], but basal ( i.e. ,\nnonpulsatile) LH secretion was reduced ( P  < 0.05;  Fig. 3B and 3C) . NK3Ra had no effect on secretory\nmass of LH per pulse ( Fig. 3D)  and total amount\nof LH secreted in a pulsatile manner ( Fig. 3E) .\nA slight increase in the orderliness ( i.e.,  decrease in the\napproximate entropy) of LH secretory pattern with the NK3Ra approached statistical\nsignificance ( P  = 0.054;  Fig.\n3F) .\nReproductive hormone response in premenopausal women in the control and\nNK3Ra-treated cycles (n = 13). Mean serum (A) LH level during single timepoint\nsampling and (B) during 8 hours of frequent sampling every 10 minutes on day 6\nto 7 of NK3Ra administration. (C) FSH level during single timepoint sampling\nand (D) during 8 hours of frequent sampling every hour on day 6 to 7 of NK3Ra\nadministration. (E) Estradiol, (F) inhibin B secretion, and (G) urinary\nprogesterone/creatinine ratio in premenopausal women (n = 11) adjusted to LH\nsurge onset day 0 with and without NK3Ra was compared in premenopausal women by\nrepeated measure two-way ANOVA at time points when paired data were available,\nfollowed by Bonferroni multiple comparisons  post hoc  analysis.\nData are given as mean ± SEM. * P  < 0.05;\n** P  < 0.01;\n*** P  < 0.001;\n**** P  < 0.0001.\nPulsatile LH secretion in premenopausal women in the control and NK3Ra-treated\ncycles (n = 8). (A) Illustrative LH pulse profile from one participant who\nunderwent blood sampling every 10 minutes for LH for 8 hours with no NK3Ra\n(closed circles) and on day 7 of NK3Ra treatment (open squares). Mean (B) LH\npulse frequency, (C) basal (nonpulsatile) LH secretion, (D) mass of LH per\npulse, (E) pulsatile LH secretion, and (F) relative orderliness or regularity\nof LH secretory pattern were compared between the control and NK3Ra-treated\ncycles. Data are given as mean ± SEM. * P \n< 0.05. ApEn, approximate entropy.\nAfter discontinuation of MLE4901 treatment, an LH surge was detected on cycle day 22\n± 1 vs 15 ± 1 in control cycles ( P  = 0.0006;  Fig. 2A and Fig. 4A) . There was no effect on the magnitude of the peak of midcycle LH secretion (23.4\n± 4.8 IU/L vs 19.7 ± 3.2 IU/L in control cycles; ns).\nSummary of the delay in timing of key events in female reproduction with the\nNK3Ra compared with no-treatment control cycles in premenopausal women.\nIndividual response of the participants (n = 13) to NK3Ra showing the day of\nthe menstrual cycle for (A) LH surge, (B) peak estradiol, (C) appearance of the\nlargest diameter of the preovulatory follicle, (D) midluteal urinary\nprogesterone (day of LH surge +7; n = 11), and (E) the length of menstrual\ncycle. ** P  < 0.01;\n*** P  < 0.001.\nFSH secretion was unchanged with the NK3Ra by single-day sampling ( Fig. 2C) . To detect subtle changes in hormone\nsecretion over time potentially overlooked by performing single-time spot blood\nsampling, a more detailed analysis of FSH secretion every hour for 8 hours after\nNK3Ra dosing showed higher FSH concentrations throughout the 8-hours during NK3Ra\nadministration compared with control cycles (ANOVA  P  < 0.03\nand  P  < 0.05 for NK3Ra cycle vs control cycles at every hour;\n Fig. 2D) .\nSerum concentrations of estradiol were affected by NK3Ra administration ( Fig. 2E) . Estradiol concentrations were\nsignificantly lower at each time point throughout treatment ( P \n< 0.05 vs control cycles) and, at the end of NK3Ra administration, estradiol\nconcentrations were markedly lower than in control cycles (166 ± 29 pmol/L vs\n446 ± 86 pmol/L, respectively, on day 12;  P  < 0.0001),\nremaining comparable to baseline levels on cycle day 5 (166 ± 29 pmol/L at end\nof treatment vs 122 ± 14 pmol/L on day 5; ns). Over the days after\ndiscontinuation of NK3Ra treatment, estradiol concentrations rose, reaching\npreovulatory levels comparable those in control cycles (690 ± 68 pmol/L vs 699\n± 62 pmol/L, respectively; ns) but 7 days later (cycle day 21 ± 1 vs 14\n± 1, respectively;  P  = 0.002;  Fig. 4B) . Inhibin B concentrations were slightly reduced during NK3Ra\nadministration, but this did not reach statistical significance ( Fig. 2F) .\nFollicle growth was suppressed in NK3Ra treatment cycles, matching the effects on\nestradiol secretion.  Figure 5A  shows follicle\ngrowth until ovulation for each woman in the control and NK3Ra-treated cycles; the\nmean data are shown in  Fig. 5B . Although there\nwas a progressive rise in diameter of the leading follicle in control cycles, this\ndid not occur during the 7 days of NK3Ra treatment ( P  = 0.0003). The\ndiameter of the leading follicle was significantly smaller than in control cycles at\neach time point throughout the 7 days of treatment ( P  < 0.05\nvs pretreatment) and at the end of treatment with the NK3Ra [ i.e. ,\non cycle day 12 (9.3 ± 0.4 mm vs 15.1 ± 0.9 mm;  P \n< 0.0001;  Fig. 5B )]. After treatment,\nnormal follicle growth resumed, reaching the same preovulatory follicle size as in\ncontrol cycles (16.1 ± 0.7 mm with NK3Ra vs 17.2±0.7 mm in control\ncycles; ns) but later (cycle day 21 ± 1 vs 15 ± 1;  P  =\n0.002;  Fig. 4C) . Consistent with the delay in\nthe timing of the LH surge, NK3Ra treatment delayed the ultrasound-determined day of\novulation in 11 of 13 subjects ( Fig. 5C) .\nFollicle and endometrial development in 13 premenopausal women in the control\nand NK3Ra-treated cycles until ovulation. (A) Follicle growth in each of the 13\npremenopausal women showed delayed development with the NK3Ra treatment (open\nsquares) compared with the control cycle (closed circles). (B) Mean follicle\ndiameter in the control and NK3Ra-treated cycle. Data include all cycles in\nwhich the leading follicle remained ( i.e. , data only include\nwomen who had not yet ovulated at the later time points). (C) Table showing the\ntime point in the control and NK3Ra cycle at which the leading follicle was no\nlonger identified at transvaginal ultrasonography. (D) Endometrial development\nin premenopausal women in control and NK3Ra-treated cycles. Data are given as\nmean ± SEM. * P  < 0.05;\n**** P  < 0.0001.\nEndometrial development was also affected by NK3Ra treatment ( P \n< 0.0001); the endometrium was significantly thinner than in control cycles at\nthe end of treatment (4.6 ± 0.4 mm vs 7.7 ± 0.5 mm, respectively, on\nday 12;  P  < 0.0001;  Fig.\n5D) . Thereafter, endometrial thickness increased, reaching a similar\nthickness to that in control cycles at the time of ovulation (8.9 ± 0.6 mm vs\n9.5 ± 1.0 mm, respectively; ns).\nConsistent with the demonstration of delayed ovulation, NK3Ra treatment delayed the\ncycle day of peak midluteal progesterone (cycle day 30 ± 2 vs 22 ± 1;\n P  = 0.002;  Fig. 4D) .\nHowever, when standardized against the day of the LH surge, luteal function was not\naffected by the NK3Ra (urinary progesterone level, 58±10 pmol/mol vs\n48±7 pmol/mol creatinine on LH surge day +7; ns;  Fig. 2G) . Menstrual cycle length was prolonged, on average, by 6 days in\nthe NK3Ra treatment cycle (35 ± 1 days vs 29 ± 1 days;\n P  = 0.0003;  Fig. 4E) .\nMLE4901 was well tolerated with no treatment discontinuations. Hematology and\nbiochemistry (including liver function) safety parameters remained stable in all\nparticipants throughout the study ( Supplemental Table 2 ). No participants had\nelevated liver function parameters related to drug administration. One woman’s\nbilirubin level was 1.4-fold higher than the upper range of laboratory reference\npretreatment (within the limits of protocol allowance) with no change during\ntreatment or thereafter. All participants returned to their usual menstrual cycle\nlength after NK3Ra treatment.\n\nThis study investigated the role of the NKB pathway in regulating physiological follicle\ndevelopment and its hypothalamic regulation through the modulation of pulsatile GnRH and\nLH secretion in healthy women. This period of the cycle includes emergence of the\ndominant follicle and its growth toward ovulation, and thus is critical for female\ncyclicity and fertility. NK3R antagonism for 7 days in the early follicular phase in\nhealthy women suppressed follicle growth and estradiol secretion, and delayed ovulation\nby the duration of treatment. NK3Ra suppressed basal LH secretion without a change in\npulse frequency, detected by frequent blood sampling following drug administration, but\nLH concentration by once-daily sampling 12 hours after the previous NK3Ra dose was\nunchanged. The half-life of MLE4901 is approximately 8.5 hours ( 26 ). FSH secretion was increased during treatment, likely reflecting\nthe reduced estradiol concentrations. FSH secretion is also promoted (relative to LH) by\nbasal rather than pulsatile GnRH secretion ( 27 ),\nthus two mechanisms may contribute to the raised FSH concentrations observed with the\nNK3Ra. These findings demonstrate that selective blockade of NK3R regulates ovarian\nfunction by reducing basal GnRH and LH secretion in the midfollicular phase, and this\neffect persisted for the duration of treatment. This confirms an important role of NKB\nin human reproduction and, furthermore, provides evidence for NKB-NK3R signaling in the\nphysiological regulation of normal follicle development in women.\nA striking finding was that the effects of the NK3Ra were reversible after\ndiscontinuation of treatment, with normal follicular estradiol production and growth\nresuming, resulting in a normal LH surge and midluteal progesterone rise, all of which\nwere delayed by the duration of treatment. Thus, although basal LH secretion was reduced\nduring treatment, there remained sufficient gonadotropin support to the emerging\ndominant follicle to prevent atresia, although whether oocyte quality might have been\ncompromised is unclear.\nThe decreased estradiol secretion for the duration of the 7 days of treatment was\nbiologically relevant, as shown by lack of endometrial development during NK3Ra\ntreatment, with subsequent growth to normal preovulatory thickness as follicle estradiol\nproduction increased after drug discontinuation. A different NK3Ra (ESN364) administered\nto healthy women for 21 days throughout the follicular phase did not result in any\nsignificant suppression in follicle growth, despite other findings being consistent with\nour data, particularly a delayed LH surge in some women and prolongation of menstrual\ncycle length ( 22 ). It is likely that the more\nvariable effect in that study between different women precluded clear demonstration of\nan effect on folliculogenesis. Although these data show that dominant follicle\ndevelopment was delayed by treatment with the NK3Ra, there was no clear evidence of an\neffect on the growth of smaller follicles, as indicated by nonsignificant changes in\ninhibin B levels.\nThe recovery of LH and FSH concentrations 12 hours after dosing may indicate that the\ndose and regimen used in this study are at the bottom of the dose-response curve.\nSimilarly, ovulation was delayed in 11 of the 13 women, and two women were\nnonresponders. It is possible that with higher doses of NK3Ra and longer duration of\nexposure, a more marked inhibitory action on LH secretion would be observed, potentially\nleading to follicle atresia and a decline in estradiol concentrations. Conversely,\nalthough follicle growth was arrested during 7 days of NK3Ra treatment, it remains\nunclear if such effect would persist with longer use. Nevertheless, consistent with\nsuppressive effects on basal LH secretion during frequent blood sampling, antagonism of\nNK3R resulted in marked ovarian effects: Follicle maturation and ovarian hormone\nsecretion were delayed, postponing the LH surge and subsequent luteal progesterone rise.\nThe lower serum estradiol levels observed with the NK3Ra treatment suggests that LH\ndependent secretion of thecal androgens was suppressed. Granulosa cell proliferation\n(and, thus, follicle growth) is predominantly driven by FSH, which was not suppressed\nbut rather increased by the NK3Ra. It may be that the NK3Ra, therefore, impaired\nfollicle function through direct intrafollicular mechanisms, because NK3R has been\nlocalized to human granulosa cells ( 28 – 30 ). The results of this\nstudy, however, are in contrast to the suppressive effects of GnRH antagonists, and of\nthe kisspeptin analog TAK448 ( 31 ), which, like\nGnRH agonists, resulted in an initial stimulation of LH secretion followed by\nsuppression. Estradiol levels remained >100 pmol/L in this study. This alleviates\nconcerns of unwanted menopausal-like adverse effects, which are associated with the use\nof GnRH analogs ( 32 ). Selective blockage of NKB\nsignaling, therefore, might have therapeutic potential in the management of sex-steroid\ndependent disorders, such as endometriosis, fibroids, and heavy menstrual bleeding, and\npotentially as a nonsteroidal contraceptive.\nThis study investigated in detail a hypothalamic mode of action of NKB during the\nfollicular phase of the menstrual cycle in healthy women, showing that, in the\nfollicular phase, NK3Ra reduced basal LH levels ( i.e. , the amount of LH\nsecreted between the pulses) and, by inference, GnRH secretion, with no effect on pulse\nfrequency. This indicates that basal rather than pulsatile LH secretion may be more\nimportant in supporting midfollicular phase follicle growth, with pulsatile secretion\nand frequency becoming more important in the lead up to ovulation, as occurs\nphysiologically ( 33 ). We have recently\ndemonstrated that the dose of MLE4901 used in this study reduced LH pulse frequency and\nabolished the correlation between estradiol and LH response to kisspeptin in a model of\nestrogen-induced LH secretion ( 20 ), which is in\ncontrast to the absence of effect on LH pulse frequency observed in the current study in\nwomen with lower estradiol during early follicular phase. This suggests that estrogen\nfeedback may have a role in modulating NKB effects on GnRH and LH secretion. The absence\nof effect on LH pulse frequency shown in the current study may reflect lower estradiol\nfeedback ( 25 ), but may also reflect differences\nin the settings of the pathways driving GnRH pulse frequency in the early follicular\nphase. Reduction in the frequency of LH and GnRH pulses by the NK3Ra has been\ndemonstrated previously in states of high LH output, in women with PCOS ( 21 ) and in gonadectomized ewes and male monkeys\n( 16 ,  19 ). Patients with inactivating mutations in the NKB pathway had diminished LH\npulse frequency ( 34 ); therefore, it remains\npossible that with larger doses of the NK3Ra, such an effect would be observed in\nhealthy women in the follicular phase of the menstrual cycle. This study may also be\nunderpowered to detect changes in LH pulse frequency.\nIn summary, NK3R antagonism in healthy premenopausal women in the early and\nmidfollicular phase of the menstrual cycle resulted in reduced LH secretion, prevented\nfollicle growth and rising estradiol secretion, and delayed ovulation by the duration of\ntreatment. Those effects were reversible after cessation of drug administration, with\nevidence of normal ovulation and luteal function. These data confirm the involvement of\nNKB in the physiological neuroendocrine control of female reproduction, with potential\ntherapeutic application in the management of sex steroid–dependent disorders.","source_license":"CC-BY-4.0","license_restricted":false}