The primary objective was to assess the efficacy of triptorelin pamoate PR 3-month in Chinese women with endometriosis by demonstrating the noninferiority of this formulation (15 mg) injected once compared with triptorelin acetate PR 1-month (3.75 mg) injected three times consecutively. The primary outcome measure was the proportion of patients who were chemically castrated (defined as E2 concentrations ≤ 184 pmol/L or 50 pg/mL) at week 12.

Efficacy was also assessed using the following secondary endpoints: percentage of patients chemically castrated at week 4 and week 8, and at weeks 4, 8, and 12 using an alternative definition of chemical castration (E2 ≤ 110 pmol/L or 30 pg/mL); change from baseline in endometriosis-associated pelvic pain (by 10 cm visual analog scale [VAS]) at weeks 4, 8, and 12; E2, LH, and FSH concentrations at weeks 4, 8, and 12; and time to menses recovery.

Exploratory endpoints included the percentage of patients chemically castrated according to the abovementioned definitions at week 24; change from baseline in endometriosis-associated pelvic pain at weeks 16, 20, and 24, and at end of study (EOS); and LH and FSH concentrations at week 24.

Pharmacokinetic (PK) parameters were also assessed for the triptorelin pamoate PR 3-month and triptorelin acetate PR 1-month formulations.

This was a phase 3, randomized, open-label, parallel group, active-controlled study conducted at 24 centers in China (NCT03232281; study funded by Ipsen). Patients were randomized 1:1 (allocation through an interactive web response system) to receive a total of two intramuscular injections with triptorelin pamoate PR 3-month (15 mg, once every 12 weeks) or a total of six intramuscular injections with triptorelin acetate PR 1-month (3.75 mg, once every 4 weeks) (Fig. 1).

Patient disposition. AE adverse event, FAS full analysis set, NCA noncompartmental analysis, PK pharmacokinetic, PP per protocol, PR prolonged-release

The study comprised a screening period of up to 5 weeks, a 24-week treatment period, and a follow-up period up to week 40 or until the recovery of menses, whichever occurred first. The visit and dosing schedules are shown in Fig. S1. The first dose was given at baseline (day 1), which occurred in the patient’s follicular phase (first to fifth day of menses). After week 12, add-back treatment (recommended as, but not limited to, tibolone 2.5 mg once daily) was administered if required, based on the investigator’s judgement. After week 24, patients were followed-up once every 4 weeks via telephone until menses recovery or week 40, whichever occurred first. After this time point, all patients were requested to attend the study site for an EOS visit.

Sparse PK samples were collected in all patients; a full PK analysis was performed for 14 patients per group, with additional blood samples collected at weeks 1, 2, 3, and 32 for analysis of E2, LH, and FSH (further details on PK analyses are provided in the Supplementary Materials).

The study was conducted according to the Declaration of Helsinki of the World Medical Association and in compliance with applicable local regulations. The protocol, amendments, and informed consent forms were approved by the independent ethics committee and institutional review boards (details of ethics committees for each study site are provided in Table S1). The ethics committee at the leading site was the Clinical Trial Ethics Committee at Peking Union Medical College Hospital, Chinese Academy of Medical Sciences. All patients provided written informed consent before study entry.

Eligible patients included women aged 18–45 years with a diagnosis of endometriosis, confirmed by laparoscopy or laparotomy within 5 years, and a history of regular menstrual cycles (21–35 days), and who were considered by the investigator to require treatment with a GnRH agonist for a period of 6 months. Patients enrolled in this study were not categorized by their stage and score of endometriosis or recurrence rate of symptoms before starting treatment. Instead, all patients with a diagnosis of endometriosis were included, with all randomized patients having received surgery for endometriosis prior to study entry, and both treatment formulations were well balanced with respect to gynecological history.

Patients were not allowed to enter the study if they were menopausal, pregnant, or lactating, or if they had received treatment with a GnRH agonist within the previous 6 months, other hormonal treatment within the previous 3 months, or a traditional Chinese medicine within the previous month.

Safety and tolerability were assessed as a secondary study objective by means of incidence and severity of adverse events (AEs), laboratory tests (biochemistry, hematology, and urinalysis), and electrocardiogram findings.

A PK comparison of the two formulations was also included as a secondary objective. Further analysis of the PK and pharmacodynamic (PD) properties of the 3-month formulation was performed in a subset of patients as an exploratory objective with E2, LH, and FSH as PD markers.

Using PK/PD subgroup data, a population PK model was built and used to derive individual PK parameters for both formulations for all patients (further details on the population PK and PD model analysis are provided in the Supplementary Materials).

A sample size of 300 (150 patients per group) was planned to ensure adequate precision and confidence in outcome estimates, based on previous study data [19]. Assuming that the proportion of patients chemically castrated at week 12 with triptorelin acetate PR 1-month would be no less than 92%, a sample size of 133 patients per treatment group was estimated to provide 85% power to demonstrate the noninferiority objective, with participant attrition estimated as 10%.

Analysis of efficacy objectives was conducted on the full analysis set (FAS), including all randomized patients receiving at least one treatment dose, and with at least one baseline and post-baseline primary endpoint assessment. The primary efficacy analysis was performed on the per-protocol (PP) set, including all patients in the FAS with a primary endpoint measurement at week 12 and without major protocol deviations. Safety endpoints were analyzed on the safety set, including all patients receiving at least one treatment dose.

The full PK profile analysis set included patients in the full PK/PD subgroup receiving at least one treatment dose, with no major protocol deviations, and with sufficient PK concentration measurements to estimate the main PK parameters: maximum concentration over a dosing interval (Cmax), time to Cmax (tmax), and area under the curve over a dosing interval (AUCtau), when applicable. The sparse PK sampling analysis set included all patients receiving at least one treatment dose, with no major protocol deviations and with at least one valid plasma concentration. The PD analysis set included all patients in the full PK/PD subgroup with sufficient PD measurements. The PK/PD relationship set included all patients receiving at least one treatment dose, with at least one valid plasma triptorelin concentration and at least one PD measurement.

The null hypothesis for the primary efficacy analysis was that triptorelin pamoate PR 3-month would be noninferior to triptorelin acetate PR 1-month when the prespecified noninferiority margin was −10%. The difference between groups in percentage of patients chemically castrated at week 12 and its two-sided 95% confidence interval (CI) were calculated using the Miettinen–Nurminen method, stratified by endometriotic surgical history and severity of endometriosis-associated pelvic pain at baseline with sample-size weighting. If the lower limit of the 95% CI was above −10%, triptorelin pamoate PR 3-month noninferiority to triptorelin acetate PR 1-month was confirmed. A sensitivity analysis was performed on the same analyses adding treatment center in the above Miettinen–Nurminen method. Supported analyses were performed using the Miettinen–Nurminen method without stratification factors.

Changes in endometriosis-associated pelvic pain and serum E2, LH, and FSH concentrations were presented using summary statistics; 95% CIs for the difference in mean values and change from baseline were calculated using a linear model for repeated measurements, adjusting for treatment group and its interaction, with visit and randomization as stratification factors. Median time to menses recovery, 25th and 75th quartile times, and 95% CIs were estimated using the Kaplan–Meier method.

Patient demographics and baseline characteristics, and safety endpoints were analyzed using descriptive statistics.

Statistical evaluations were performed using Statistical Analysis System (SAS)® software (version 9.4). The noncompartmental analysis was performed using SAS software on the PK data collected for the PK/PD subgroup to assess the individual PK parameters for both formulations. Modeling was performed on PK data from all patients using the nonlinear mixed effects model (further details of PK analyses are provided in the Supplementary Materials).