Tolerability and safety of the estetrol/drospirenone combined oral contraceptive: Pooled analysis of two multicenter, open-label phase 3 trials ✩ , ✩✩

Objectives: To evaluate tolerability and safety of estetrol (E4) 15 mg/drospirenone (DRSP) 3 mg oral con- traceptive using pooled data from two, multicenter, phase 3 trials. Study design: The two trials enrolled participants aged 16 − 50 years with a body mass index ≤ 35.0 kg/m 2 to use E4/DRSP in a 24/4-day regimen for up to 13 cycles. We pooled data from participants who used at least one E4/DRSP dose and had a follow-up assessment to analyze adverse events (AEs), vital signs, and laboratory parameters, including serum lipids, glucose, glycated hemoglobin, and potassium. We consolidated similar Medical Dictionary for Regulatory Activities preferred terms into groupings. Results: Of 3725 participants enrolled, we included 3417 in the analyses of whom 1786 (52.3%) reported ≥ 1 AE. Most participants with reported AEs had AEs that investigators rated as mild or moderate ( n = 1665, 93.2%); of participants reporting AEs, 1105 (61.9%) did so during cycles 1 to 3. In total, 981 (28.7%) participants experienced ≥ 1 treatment-related AE, most frequently related to bleeding complaints ( n = 323, 9.5%), breast pain or tenderness ( n = 136, 4.0%), acne ( n = 113, 3.3%), and mood disturbance ( n = 111, 3.2%). Discontinuation due to treatment-related AEs occurred in 272 participants (8.0%), with only bleeding complaints ( n = 97, 2.8%) and mood disturbance ( n = 38, 1.1%) at rates exceeding 1%. Three participants experienced serious AEs, which the site investigators considered treatment-related: one ve- nous thromboembolism, one worsening of depression, and one ectopic pregnancy. We found no clinically relevant changes in weight, blood pressure, heart rate, or laboratory parameters during treatment. Conclusions: E4/DRSP is associated with a favorable tolerability and safety proﬁle. Implications statement: Pooling data allowed for a robust assessment of tolerability and safety, includ- ing relatively infrequent events. Other than bleeding complaints and mood disturbance, no adverse event resulted in E4/DRSP discontinuation at rates > 1%. Post-marketing surveillance studies are needed to eval- uate long-term safety of the E4/DRSP COC and population-based venous thromboembolism risks.


Introduction
Combined oral contraceptives (COCs) are the most commonly used method of reversible contraception in North America and Europe [ 1 , 2 ], with most COCs comprised of ethinyl estradiol (EE) in combination with a progestin [3] . New COC formulations introduced over the past few decades have aimed to cause fewer side effects while maintaining efficacy; still, some users experience side-effects, including sexual dysfunction, mood changes, weight gain, breast tenderness, and unscheduled uterine bleeding [4][5][6][7][8] . Importantly, currently available COCs are associated with rare but serious cardiovascular adverse effects including venous thromboembolism (VTE) [9][10][11] . The adverse effects of COCs can be a barrier to use and result in discontinuation, with the potential for unintended pregnancies [12][13][14][15] . Therefore, a new COC with a favorable tolerability and safety profile could provide a beneficial option.
Estetrol (E4) is the estrogenic component of a new COC formulated with the progestin drospirenone (DRSP). E4 is naturally produced by the human fetal liver and is synthesized from a plant source for clinical use. This native estrogen has properties distinct from other natural and synthetic estrogens, displaying tissue-selective agonistic and/or antagonistic estrogenic properties elicited through selective nuclear estrogen-receptor (ER) α activation, but not membrane ER α activation in several tissues including the breast [16][17][18][19] . The selective receptor activity may result in a limited impact on hemostasis parameters, breast tissue, endocrine parameters, liver proteins, lipid profiles, and carbohydrate metabolism, while sustaining endometrial proliferation [20][21][22][23][24] . E4 15 mg (as monohydrate, equivalent to anhydrate 14.2 mg)/DRSP 3 mg has recently been approved in the United States (US), Canada, the European Union (EU), and Australia, with marketing authorization supported by the efficacy and safety results of two phase 3 studies [ 25 , 26 ]. We performed a pooled analysis of adverse events (AEs), laboratory data, and vital signs across the two phase 3 studies to further characterize the tolerability and safety profile of E4/DRSP in a larger spectrum of individuals.

Materials and methods
Investigators enrolled participants into two parallel phase 3 clinical trials from June 2016 through April 2018 (Europe/Russia) and from August 2016 through November 2018 (US/Canada). For this analysis, we included data from participants who had confirmed use of at least one dose of study drug and a follow-up visit and/or call. We analyzed AEs and serious AEs (SAEs), overall and treatment-related, by severity and cycle, together with vital signs and laboratory parameter abnormalities. In addition, we calculated the proportion of affected cycles against total number of cycles for treatment-related AEs.
The methods and outcomes of the individual trials have been previously reported [ 25 , 26 ]. Briefly, investigators enrolled healthy, heterosexually active, pre-menopausal participants (18 −50 years Europe/Russia trial; 16 −50 years US/Canada trial) with a body mass index (BMI) ≤35.0 kg/m 2 , and a history of regular menstrual cycles (21 −35 days) when not on hormonal contraception. Investigators excluded individuals with contraindications to COC use based on World Health Organization (WHO) medical eligibility criteria [27] . Specific exclusion criteria included a history of thromboembolic, cardiovascular or cerebrovascular disorder, hypertension (systolic blood pressure [SBP] ≥140 mmHg or diastolic blood pressure [DBP] ≥90 mmHg) and use of any nicotine-containing products for persons ≥35 years old. Participants received the study drug in a blister pack containing 24 active E4/DRSP tablets and 4 inactive tablets, to be taken once-daily for 28 days for up to thirteen cycles.
Study staff planned four follow-up study visits on-treatment (Cycles 2, 4, 7, and 10) and one at the end of treatment (Cycle 13 or early discontinuation). Participants used paper diaries to record medication intake, other contraceptive methods used, sexual activity, vaginal bleeding and/or spotting events, and AEs. At each visit, study staff reviewed the diaries, collected used study drug packets, dispensed new drug, and asked participants about any changes in medical conditions, other medication use, and the occurrence of AEs. Investigators assessed study drug compliance based on diary entries per 28-day cycle and counted any day with a missing entry as no pill intake. For the analysis, we assessed treatment compliance as the reported number of pills taken divided by the expected number of pills taken based on duration of participation.
Investigators evaluated the frequency and severity of AEs, including clinically relevant changes or abnormalities in routine laboratory parameters or physical examination findings. Investigators assessed clinical laboratory parameters (including hematology, serum chemistry, and lipid profiles) at Screening, Cycle 7, and Cycle 13, and vital signs (SBP, DBP, heart rate, and weight) at Screening and Cycles 2, 4, 7, 10, and 13. For each AE, site investigators determined whether the AE should be categorized as an SAE and assessed the relationship of the AE to the study drug as treatment-related or not. The investigators determined that the AE was probably or possibly related to the study drug if there was a reasonable time relationship to the study drug intake and if it was unlikely to be due to an underlying illness or concurrent treatment. Site investigators recorded the intensity of each AE as mild (transient and well-tolerated by the study subject), moderate (temporary interference with daily living), or severe (substantially interfered with daily living to the point of being incapacitating and/or life-threatening). For the analysis, we classified AEs using version 20.0 of the Medical Dictionary for Regulatory Activities (MedDRA) preferred terms and consolidated similar preferred terms into groupings as presented in Supplemental Table 1. To calculate the number of affected cycles, we compared the beginning and end date of each treatment-related AE with the beginning and end date of each cycle. If a participant reported the event more than once within a single cycle, we counted it only once for the cycle. The proportion of affected cycles was calculated as the ratio of the number of impacted cycles and the total number of cycles of all participants during the study.

Participants and compliance
Overall, we enrolled 3725 participants; 93 participants discontinued before study drug initiation, and we could not confirm intake in 215 participants as they had no follow-up contact, leaving 3417 participants in this pooled analysis ( Fig. 1 ). The baseline characteristics of the participants are presented in Table 1 . By the end of Cycle 3, 6, 9, and 13, 466 (13.6%), 771 (22.6%), 1035 (30.3%), and 1183 (34.6%) participants had discontinued, respectively. The disposition of participants through the two clinical trials is shown in Figure 1 ; the most common reasons for discontinuation were loss to follow-up ( n = 328 [9.6%]) and consent withdrawal ( n = 261 [7.6%]).

Serious adverse events
Thirty-eight (1.1%) participants experienced an SAE, of which 6 (0.2%) discontinued the study related to the event. Investigators assessed 3 (0.1%) as treatment-related: one hospitalization for worsening of depression (no discontinuation of study drug), one ectopic pregnancy, and one lower extremity VTE. The VTE event resolved without sequelae after anticoagulant treatment. One death occurred, related to a self-administered fentanyl and alprazolam overdose, which the investigator assessed as unlikely related to the study drug.

Body weight, vital signs and laboratory parameters
We did not observe clinically significant changes from baseline in mean values of SBP, DBP, and heart rate during treatment (Cycle 7 and Cycle 13) and measured changes in body weight from baseline to end of treatment were minimal (Supplemental Figure 1). Overall, 111 (3.3%) participants had an SBP of ≥140 mmHg and/or a DBP ≥90 mmHg during the treatment period. In addition, 745 (21.8%) participants gained ≥5% of their baseline weight and 483 (14.1%) lost ≥5% of their baseline weight during treatment. Eleven (0.3%) participants experienced hypertension or increased blood pressure as an AE, of which investigators considered 6 (0.2%) to be related to the study drug and two discontinued for the event (Supplemental Table 2).
We also did not observe clinically significant mean changes from baseline in serum lipids, glucose, glycated hemoglobin (HbA1c), and potassium during treatment (Supplemental Figure  1). Investigators reported hyperkalemia/increased blood potassium as an AE in 7 (0.2%) participants (Supplemental Table 2) which included one participant with a value in the normal range (5.2 mmol/L, normal 3.5-5.3 mmol/L); none had any associated symptoms. One participant, with a potassium of 4.4 mmol/L at baseline, had a value of 8.0 mmol/L 18 days after last study drug use. The other 5 participants had potassium values of 5.5 to 6.0 mmol/L, one of whom (6.0 mmol/L) discontinued.  Table 1 for the MedDRA preferred terms of the groupings.  Table 1 for the MedDRA preferred terms of the groupings.

Table 3
Proportion of cycles with treatment-related AEs in participants using estetrol/drospirenone for up to 13 cycles ( N = 3417) a .

Discussion
We demonstrated that most E4/DRSP users experienced high tolerability with a favorable safety profile based on the combined data from two pivotal phase 3 trials involving 3417 participants with 35,093 cycles of exposure. Regulatory agencies pooled safety data for safety considerations during drug approval. We grouped similar AEs to provide more clinically meaningful information for healthcare professionals and COC users. Overall, about 29% of participants reported AEs assessed by investigators as treatmentrelated, the most common of which were typical of those selfreported for other COCs including bleeding complaints, breast pain or tenderness, acne, mood disturbance, headache, dysmenorrhea, and increased weight [ 6 , 28 ]. Increased potassium, a potential consequence of the weak potassium-sparing diuretic effect of DRSP, occurred uncommonly; this finding was reported in 7 (0.2%) asymptomatic participants.
The background annual VTE incidence for EE-containing COC users is 5 to 10/10,0 0 0 woman-years [ 29 , 30 ]. In our pooled analysis, a single participant with no known baseline risk factors experienced a lower extremity VTE. The overall estimated annual VTE incidence rate across the full E4/DRSP clinical program (pooled phase 2 and 3 trials) is 3.66/10,0 0 0 woman-years [31] . The risk of VTE is considered highest in the first 12 months of using a new COC [32] , with multiple studies reporting a small increased risk in new users compared to switchers [32][33][34] , perhaps because switchers have already demonstrated a lower risk for VTE [34] . A similar increased risk has also been observed in those re-starting a COC after an absence [35] . While nearly half of the participants in this pooled analysis were starters, that is, past contraceptive use > 3 months before initiating study drug or no prior use, only 19% of participants were true new users. A study with a larger population of true new users may have different results, although the low proportion of true new users our studies are in-line with other phase 3 COC trials [ 36 , 37 ]. Because VTE is rare with COC use, a population-based post-marketing study is needed to confirm VTE risk with the E4/DRSP formulation.
We found no clinically relevant changes with E4/DRSP use in parameters that indicate cardiovascular risk including blood pressure, heart rate, lipids, glucose, HbA1c, and potassium. Clinical studies have consistently demonstrated that E4 has a limited effect on lipids, hemostasis, and carbohydrate metabolism [ 20 , 22 , 24 , 38 ]. However, these trials of E4/DRSP use excluded participants with major cardiovascular risk factors in accordance with WHO eligibility criteria for COCs [27] . Until further data are available, the contraindications related to cardiovascular risk should remain the same as other COCs [ 39 , 40 ].
A strength of this analysis is that we used pooled data from the two pivotal trials, allowing an evaluation of the risks and benefits of E4/DRSP in a large number of participants. Pooling data allowed for a robust assessment of tolerability and safety, including relatively infrequent events. In addition, the trials, conducted in North America, Europe, and Russia, included participants with a range of different demographic factors. However, some limitations are worth noting. Most participants were white (82.9%), and only 15% were obese. Therefore, the results of the pooled analysis may not be generalizable to populations with different characteristics. Furthermore, as is standard for pivotal contraceptive regulatory trials, the study did not include a comparator contraceptive, and therefore we cannot provide any direct comparison with other COCs or methods.
In conclusion, this pooled analysis adds to the body of evidence supporting E4 in combination with DRSP as a COC with a favorable tolerability and safety profile. Post-marketing surveillance studies will provide additional data to evaluate the long-term safety of the E4/DRSP COC. bers of the Estelle Scientific Advisory Boards for their valuable advice. Amanda Prowse, Patricia de Groot and Lily Fitzgerald at Terminal 4 Communications, Hilversum, the Netherlands provided medical writing support. PRA Health Sciences acted as CRO for the clinical trials. Maud Hennion, Cédric Charlot and Fabrice Nollevaux (Pharmalex, Belgium) provided statistical support for this pooled analysis.

Supplementary materials
Supplementary material associated with this article can be found, in the online version, at doi: 10.1016/j.contraception.2022.10. 004 .