Neurokinin 3 receptor antagonism as a novel treatment for menopausal hot flushes: a phase 2, randomised, double-blind, placebo-controlled trial

Abstract

Background: Hot flushes affect 70% of menopausal women and often severely impact physical, psychosocial, sexual, and overall wellbeing. Hormone replacement therapy is effective but is not without risk. Neurokinin B signalling is increased in menopausal women, and has been implicated as an important mediator of hot flushes.

Methods: This phase 2, randomised, double-blind, placebo-controlled, single-centre, crossover trial assessed the effectiveness of an oral neurokinin 3 receptor antagonist (MLE4901) on menopausal hot flushes. Eligible participants were healthy women aged 40-62 years, having seven or more hot flushes in every 24 h of which some were reported as being severe or bothersome, who had not had a menstrual period for at least 12 months, and who had not been taking any medication shown to improve menopausal flushes in the preceding 8 weeks. Participants received 4 weeks of MLE4901 (40 mg, orally, twice daily) and placebo (orally, twice daily) in random order separated by a 2 week washout period. Randomisation was completed by a central computer, and participants were allocated to treatment number in numerical order. The primary outcome was the total number of hot flushes during the final week of both treatment periods. Analyses were by intention to treat and per protocol using generalised linear mixed models and standard crossover analysis. All analyses were prespecified in the study protocol. The trial is registered at ClinicalTrials.gov, number NCT02668185.

Findings: 68 women were screened between Feb 3 and Oct 10, 2016, of which 37 were randomly assigned and included in an intention-to-treat analysis. 28 participants completed the trial and were included in a per-protocol analysis. MLE4901 significantly reduced the total weekly number of hot flushes by 45 percentage points (95% CI 22-67) compared with the placebo (intention-to-treat adjusted means: placebo 49·01 [95% CI 40·81-58·56] vs MLE4901 19·35 [15·99-23·42]; adjusted estimate of difference 29·66 [17·39-42·87], p<0·0001). Treatment was well tolerated. Three participants developed a transaminase rise (alanine aminotransferase 4·5-5·9 times the upper limit of normal) with a normal bilirubin 28 days after starting MLE4901, which normalised within 90 days.

Interpretation: Treatment with a neurokinin 3 receptor antagonist (MLE4901) could be practice changing as it safely and effectively relieves hot flush symptoms without the need for oestrogen exposure. Larger scale studies of longer duration are now indicated.

Funding: UK Medical Research Council and National Institute for Health Research.

Figure 1

Summary of protocol Baseline period: participants underwent a 2 week period to gather baseline data on hot flush frequency, severity, bother, and interference (Hot Flash Related Daily Interference Scale). If the inclusion criteria regarding hot flush frequency and severity were met at the end of this period then they were assigned to the active phase of the study. Intervention 1 (double-blind): all participants randomly assigned to either 4 weeks of treatment with oral, twice daily 40 mg MLE4901 or exact-match placebo. Washout period: all participants underwent a 2 week washout period after intervention 1 (half-life of MLE4901 is 8·5 h). Intervention 2 (double-blind): all participants then switched to receive either 4 weeks of treatment with oral, twice daily exact-match placebo or oral, twice daily 40 mg MLE4901 depending on which intervention they received first. Monitoring period: a subsequent 2 week period to complete safety monitoring.

Figure 2

Trial profile Intention-to-treat analysis included all participants who were randomly assigned and received study medication (placebo or MLE4901; n=37). Per-protocol analysis included all participants who appropriately completed both treatment periods (n=28). See the appendix (pp 1–3) for more details about inclusion and exclusion criteria, and the number of participants affected by each.

Figure 3

Primary endpoint ITT analysis (A) Whole group ITT analysis (n=37) irrespective of treatment assignment order using adjusted means from crossover analysis with 95% CIs: percentage change in hot flush frequency (total number of hot flushes) during the final week of the 4 week treatment period with MLE4901 and placebo compared with hot flush frequency (total number of hot flushes) during the final week of the 2 week baseline period. Statistical analysis incorporated a total of seven daily counts for each of the study weeks analysed, and is based on a crossover model including treatment and period as fixed effects, subject as a random effect (within sequence), and baseline flush count as a covariate. All other possible demographic covariates were tested in the model but none were significant and therefore all were excluded from the final model. The model used is a generalised linear model with gamma error structure. Tests for sequence (order), and period, effect across all our models confirmed neither were significant. (B) Subgroup ITT analysis (n=37) by treatment assignment group using participants' unadjusted (raw) data with 95% CIs: percentage change in hot flush frequency (total number of hot flushes) during the final week of the 4 week treatment period with MLE4901 and placebo compared with hot flush frequency (total number of hot flushes) during the final week of the 2 week baseline period depending on whether the participant received MLE4901 or placebo as the first or second intervention. ITT=intention to treat.

Figure 4

Luteinising hormone pulse analysis JDV used a blinded deconvolution method with 93% sensitivity and specificity to analyse luteinising hormone pulsatility by calculating the number of luteinising hormone pulses (A), the mean amplitude of luteinising hormone pulses (B), and the orderliness of the pulses (approximate entropy; the lower the number the more ordered the pulses are, with zero denoting perfect orderliness; C). A generalised linear model was used for analysis with a Poisson error structure for number of pulses, and with a gamma structure for mean amplitude and orderliness of luteinising hormone pulses. A standard crossover analysis was implemented, with period, administration sequence, and treatment as fixed effects and subject as a random effect. Box plots: line, median; box, IQR; whiskers extend to the extremes of the data (minimum and maximum values).