Metabolic and laboratory effects of a progestin-only pill containing drospirenone 4 mg in comparison to desogestrel 75 µg: a double-blind, double-dummy, prospective, randomised study

Abstract

Purpose: Steroid hormones in hormonal contraceptives influence a variety of metabolic parameters. Mainly, ethinyloestradiol (EE) in combined hormonal contraceptives (CHC) is a potent inducer of hepatic coagulation factors, has an impact on cholesterol and triglyceride levels and glucose tolerance. Progestins in CHC modify the oestrogen effects in different ways, depending on their pharmacologic properties. The metabolic impact of progestin-only contraceptives is generally considered low.

Methods: The influence of novel drospirenone (DRSP)-only pill (4 mg DRSP in 24/4 intake) on a variety of lipid-, carbohydrate- and bone metabolic parameters and on haemostatic variables including clotting factors and D-Dimer level was evaluated in comparison to 0.075 mg desogestrel (DSG) during a multicentric, prospective, double-blind, double-dummy clinical trial with 1190 participants over nine treatment cycles.

Results: For both DRSP and DSG, there was a decrease in cholesterol (total, HDL, and LDL) and triglyceride levels. No relevant influence on glucose, insulin, and c-peptide levels or bone remodelling markers were detected in both treatment groups. Considering the coagulatory parameters, there was no impact on hemostasis.

Conclusions: The results confirm the beneficial properties of the drospirenone-only pill.

EudraCT Number: 2011-002396-42

与75μg去氧孕烯酮相比, 含4mg屈螺酮孕激素单药的代谢和实验室效应：一项双盲、双模拟、前瞻性、随机研究 摘要

目的：激素类避孕药中的类固醇激素影响多种代谢参数。主要是, 复方激素避孕药（CHC）中的炔雌醇（EE）是肝凝血因子的强效诱导剂, 对胆固醇和甘油三酯水平以及糖耐量有影响。CHC中的孕激素根据其药理学特性以不同的方式调控雌激素效应。仅应用孕激素的避孕药对代谢的影响通常被认为很低。

方法：在一项多中心、前瞻性、双模拟、双盲研究中, 对1190名参与者进行了九个治疗周期的双模拟临床试验, 评估新型屈螺酮（DRSP）单用药丸（24/4摄入4 mg DRSP）, 与0.075 mg去氧孕烯（DSG）相比, 对各种脂质、碳水化合物和骨代谢参数以及凝血变量（包括凝血因子和D-二聚体水平）的影响。

结果：DRSP和DSG均降低胆固醇（总胆固醇、高密度脂蛋白和低密度脂蛋白）和甘油三酯水平。在两个治疗组中均未检测到对葡萄糖、胰岛素、c肽水平及骨代谢标记物的相关影响。根据凝血参数, 对凝血没有影响。

结论：结果证实了屈螺酮单药的有益特性。

Keywords:

• Drospirenone-only pill
• metabolic parameters
• laboratory data

Introduction

Shortly after introducing the first combined oral contraceptives (=COC) in the 1960s, the first cases of venous thrombotic events (VTE) were reported in connection with the use of the contraceptive method [1]. Since the oestrogen component (ethinylestradiol (=EE)) of the COCs was primarily blamed as a possible cause of the onset of thrombosis and other side effects such as weight gain, bleeding disorders, and nausea, its dose was continuously reduced, indeed leading to an amelioration of the side effects [2]. EE, as the most frequent component in COC is known to exert a strong metabolic impact as it increases the synthesis of various liver proteins, including SHBG (sexual hormone-binding globulin) and CBG (corticosteroid-binding globulin), increases the production of lipoproteins, such as VLDL (very low-density lipoprotein) and HDL (high-density lipoprotein) and triglycerides levels, while decreasing LDL (low-density lipoprotein) levels and were also shown to have an impact on glucose tolerance [3–6].

Above all, EE is a potent inducer of coagulation parameters, including fibrinogen, prothrombin, and factors VII, VIII, and X. Also, Protein C activity, protein C inhibitors, factor V and other parameters are altered. Overall, the haemostaseologic balance is shifted towards a more pro-coagulatory state [7–9]. This may be reflected by the increase in the risk for VTE or ATE (arterial thromboembolic events) among COC users compared to non-users [10,11] and is independent of the route of administration [9,12].

Compared to EE, oestradiol (E2), oestradiol-valerate (E2V), or estetrol (E4) exhibit lower effects on hepatic metabolism. Therefore, replacing EE with those variants in a COC is associated with fewer metabolic effects [13–17], an improved safety profile, and a lower cardiovascular risk compared to other COCs [18–20].

The type of progestin used in COCs modulate the endocrine and metabolic effects as they differ in their pharmacological profiles and actions at different steroid hormone receptors [4,21,22]: following the first COCs with nortestosterone derivatives like lynestrenol and ethynodiol acetate, progestins including levonorgestrel (LNG), gestodene and desogestrel were developed. With progestins like clormadinone acetate, dienogest (DNG), and drospirenone, progestins with anti-androgenic activity broadened the contraceptive choice [23].

More androgenic progestins like LNG in CHC were associated with a less favourable effect on lipid metabolism compared to anti-androgenic compounds like DNG (see, e.g., 3, 5, 6 for reviews)

Progestins in COC also modulate the risk for VTE that is imposed on COC users by the oestrogen compound. Levonorgestrel, for example, can at least partially antagonise the pro-coagulatory action of EE, while other progestins like dienogest, desogestrel, or drospirenone cannot [7–9]. This may be reflected by the differences in VTE risk of COCs of different compositions as extensively described in the literature [9–11,24] and corresponding recommendations of the regulatory authorities [25].

Progestin-only contraceptives, on the other hand, do not increase the ATE/VTE risk [10,11,24,26–28] and may therefore also be recommended for patients with a broad range of cardiovascular and metabolic risk factors [20].

Recently, a novel progestin-only oral contraceptive with 4 mg drospirenone (DRSP) has been approved by the FDA and all EU countries. Its cardiovascular safety and cycle control profile has been documented in several phase 3 clinical trials [29–34].

In the present study, the influence of 4 mg DRSP on biochemical and haemostatic laboratory parameters was compared to desogestrel 75 µg in a prospective, randomised, double-blind, double-dummy trial with 1190 participants over nine treatment cycles.

Material and methods

General

This clinical phase III study was a prospective, double-blind, double dummy, randomised controlled trial including 88 centres in Austria, Czech Republic, Germany, Hungary, Poland, Romania, Slovakia, and Spain. The studies were performed between 1 August 2012, and 27 January 2014. The protocol was designed and conducted according to existing legal regulations and in accordance with good clinical practice in the conduct of clinical trials and the declaration of Helsinki, including recommendations made in the European Medicines Agency (EMA) CHMP Guideline on Clinical Investigation of Steroid Contraceptives in Women. Institutional review board approval was obtained for all study sites. The primary objective of the trial was the description of the overall pearl index. Safety and tolerability of DRSP 4 mg in comparison with desogestrel were assessed by recording adverse events and assessment of vital and laboratory data.

Ethical approval

All participants gave their written informed consent for participation in the clinical trial after obtention of the correspondent ethical committee approval. For each of the investigational centres, ethical permission was obtained. The overall approval for the trial with the leading ethical committee was given the 13.07.2012 by the Landesamt für Gesundheit und Soziales Berlin, Geschäftstelle der Ethik Kommission des Landes Berlin, number 11/0606 EK.

Study medication

The study medication was one tablet of 4 mg non-micronized DRSP per day with consecutive administration of 24 active pills and four placebo tablets with no tablet-free interval between 2 successive cycles.

DSG 0.075 mg (in a regimen of 28 active pills) was chosen as the comparator. Medication compliance was measured using an electronic diary.

Study populations

Fertile women aged between 18 and 45 years were included in the trial. Main inclusion criteria included the absence of uncontrolled current disease and a blood pressure <140mmHg (systolic) and <90 mmHg (diastolic).

Starter: First administration of a hormonal contraceptive or at least 4-month break after the administration of another hormonal contraceptive. Switcher: Direct switch from another hormonal contraceptive to the studied drug with no gap in administration. (Table 1 depicts the baseline characteristics of the participants).

Table 1. Baseline patients characteristics.

		Study population	DSG 0.075mg (N = 332)
	Statistic	DRSP 4mg (N = 858)
Age (years)	Mean (SD)	28.9 (7.1)	28.9 (7.1)
Age group
≤35 years	n (%)	682 (79.5)	259 (78.0)
>35 years	n (%)	176 (20.5)	73 (22.0)
Ethnicity
Caucasian	n (%)	856 (99.8)	331 (99.7)
BMI [kg/m^2]	Mean (SD)	22.96 (3.537)	22.82 (3.905)
Min/Max	16.6/41.0	15.9/38.0
BMI group
<30 kg/m^2	n (%)	828 (96.5)	316 (95.2)
≥30 kg/m^2	n (%)	30 (3.5)	16 (4.8)
BP group
SBP <130 and DBP <85 mmHg	n (%)	727 (84.7)	290 (87.3)
SBP ≥130 and DBP ≥85 mmHg	n (%)	131 (15.3)	42 (12.7)
Subject status
Switcher	n (%)
Direct switcher	n (%)	628 (73.2)	259 (78.0)
Indirect Switcher	n (%)	39 (4.5)	14 (4.2)
Starter	n (%)	191 (22.3)	59 (17.8)
Unknown	n (%)	–	–
VTE risk factor
Presence of at least one risk factor	n (%)	142 (16.5)	59 (17.8)
Previous delivery
Yes	n (%)	395 (46.0)	150 (45.2)
Regular menstrual bleeding during the last 6 cycles
Yes	n (%)	786 (91.6)	305 (91.9)
Prior treatment with sex hormones and modulators of the genital system
Yes	n (%)	469 (54.7)	195 (58.7)

BMI: Body mass index; DRSP: drospirenone; DSG: desogestrel; N: Number of participants.

Sample size

The sample size for the trial had been calculated with respect to the primary efficacy endpoint of the phase 3 trial (Overall pearl index), and the plausibility of the calculations has been published elsewhere [31]. Laboratory data were evaluated for the safety set of participants (all subjects, who were randomised and received at least one dose of the medication and had at least one post-baseline assessment of a safety/tolerability measurement). It comprised 858 women in the DRSP group and 332 in the reference group. From a subgroup of 68 women, particular clinical laboratory parameters were evaluated (See below).

Analysed parameters

The assessment was performed at screening (visit 1; baseline), at visit 3 (day 24 ± 2 of the 3^rd medication cycle), and visit 5 (after completion of the last medication cycle; endpoint (defined as last non-missing value for subjects prematurely terminating the trial).: Biochemical parameters (total -, HDL- and LDL cholesterol, triglycerides, albumin, total and direct bilirubin, gamma-glutamyl transferase), Thyroid function (TSH), Hematological parameters (Haemoglobin, Erythrocytes, Haemoglobin), and non-fasting plasma glucose was assessed for the complete safety set.

In a subset of 68 participants, additional parameters were evaluated (only at baseline and endpoint): Insulin, C-peptide, haemostaseological parameters (Factor VII, -VIII, protein C activity, antithrombin III activity, resistance to activated protein C [APC resistance], D-Dimer level) and bone metabolism parameters (bone alkaline phosphatase, cross-linked C-terminal telopeptides (CTX). Blood samples were collected under fasting conditions and were drawn between 8, and 9 am, to prevent an impact of the circadian rhythms and ensure the comparability of the results.

Deviations from the reference ranges were evaluated by the investigator, and clinically significant laboratory findings or worsening of relevant conditions were reported as adverse event (AE). All laboratory assessments were performed at a central laboratory (LKF; Laboratorium für klinische Forschung GmbH in Schwentinental, Germany).

Statistics

Clinical laboratory variables were summarised for each treatment group at each visit. Laboratory values were compared using a 2-sample t-test. The probes were all collected before 11:00 am.

Results

Biochemical parameters

No clinically significant changes in biochemical parameters were observed, with all values remaining in the reference range of the respective parameter (Table 2). Mean and median values of cholesterol and triglyceride levels decreased over time in both treatment groups. For triglyceride levels, the reduction was significantly stronger in the DRSP group (−0.111 mmol/L (DRSP) vs −0.226 mmol/L (DSG); p = 0.0351). The difference was considered clinically not relevant since the values at endpoint were in the same range (Mean (SD): 0.927 (0.4662) vs. 0.906 (0.4821)).

Table 2. Lipid- and carbohydrate metabolic parameters.

Parameter	Visit		DRSP 4 mg	DSG 0.075 mg	p-value
Cholesterol (mmol/L) [R.r: 4.14–6.73]	Baseline	N	514	195
		Mean (SD)	4.752 (0.7936)	4.666 (0.8663)	0.2080
	Endpoint	N	522	199
		Mean (SD)	4.473 (0.7863)	4.289 (0.8050)	0.0052
	Change	n	420	162
		Mean (SD)	−0.247 (0.8108)	−0.373(0.8018)	0.0906
Cholesterol-HDL (mmol/L) [R. r: 1.09–2.28]	Baseline	n	514	195
		Mean (SD)	1.712 (0.3909)	1.690 (0.4185)	0.5180
	Endpoint	n	522	199
		Mean (SD)	1.578 (0.3443)	1.522 (0.3528)	0.0520
	Change	n	420	162
		Mean (SD)	−0.140 (0.3414)	−0.141 (0.3516)	0.9746
Cholesterol-LDL (mmol/L) [R.r.: 1.97–5.65]	Baseline	n	514	195
		Mean (SD)	2.680 (0.7299)	2.588 (0.8104)	0.1462
	Endpoint	n	522	199
		Mean (SD)	2.566 (0.7183)	2.457 (0.7357)	0.0697
	Change	n	420	162
		Mean (SD)	−0.091 (0.7059)	−0.147 (0.7373)	0.3919
Triglycerides (mmol/L) [R.r.: 0.8–1.94]	Baseline	n	514	195
		Mean (SD)	1.064 (0.5281)	1.124 (0.6561)	0.2131
	Endpoint	N	522	199
		Mean (SD)	0.927 (0.4662)	0.906 (0.4821)	0.5910
	Change	n	420	162
		Mean (SD)	−0.111 (0.5629)	−0.226 (0.6520)	0.0351
Glucose (mmol/L) [R.r.; 4.22–6.11]	Baseline	n	335	128
		Mean (SD)	5.267 (0.7604)	5.158 (0.6318)	0.1477
	Endpoint	n	303	120
		Mean (SD)	5.211 (0.7979)	5.125 (0.8004)	0.3157
	Change	n	210	85
		Mean (SD)	−0.019 (1.0288)	−0.089 (0.8050)	0.5736
Insulin (pmol/L) [R.r.: 18–173]	Baseline	n	39	29
		Mean (SD)	111.2 (151.43)	102.8 (93.69)	0.7922
	Endpoint	n	36	26
		Mean (SD)	149.3 (225.97)	129.9 (148.22)	0.6958
	Change	n	33	28
		Mean (SD)	36.6 (178.25)	25.6 (155.89)	0.8008
C-peptide (nmol/L) [R.r.: 0.36–1.46]	Baseline	n	39	29
		Mean (SD)	0.9013 (0.46614)	0.9066 (0.32870)	0.9585
	Endpoint	n	36	28
		Mean (SD)	0.9772 (0.56526)	1.0151 (0.59616)	0.7961
	Change	n	33	28
		Mean (SD)	0.0709 (0.50457	0.1023 (0.60780	0.8260

Furthermore, standard hepatic diagnostic markers albumin, bilirubin, and Gamma-GT were evaluated. A slight increase was observed for both treatment groups in all three parameters. For albumin and bilirubin, the increase from baseline to endpoint was statistically significantly more pronounced in the group of DSG users (Mean (SD) DRSP vs DESO: Albumin: +0.1(3.30) g/L vs +0.9(3.04) g/L; p = 0.0007; and Bilirubin, direct: +0.31 (1.446) µmol/L vs +0.55(1.621) µmol/L; p = 0.0159). The changes were considered clinically not relevant, as they were minimal. There was no statistically significant difference in gamma-GT between the treatment groups.

There was no significant influence on the TSH levels in both treatment groups. Changes (Median (SD) were 0.020 (1.1707) mU/L for DRSP and −0.020 (1.2034) mU/L for DSG.

An overview of parameters of carbohydrate metabolism (C-peptide, insulin, and plasma fasting glucose) is also provided in table 2. No statistically significant differences at baseline or endpoint were observed between the treatment groups. No clinically relevant changes were observed in the mean or median values of insulin, plasma fasting glucose, or c-peptide.

Haematology

Hematological parameters were within reference ranges at baseline and endpoint (Table 3). The mean changes from baseline to endpoint were statistically significantly different between the treatment groups for erythrocyte count (mean (SD) DRSP vs DSG: −0.022 (0.2810) 10¹²/L vs 0.046 (0.065) 10¹²/L; p = 0.002), haemoglobin (−0.03 (0.555) mmol/L vs 0.09 (0.515) mmol/L; p = 0.0016) and haematocrit (0.010 (0.0298) % vs. 0.015 (0.0303) %; p = 0.043). In all cases, changes were more pronounced in the DSG group. Nevertheless, the differences were considered clinically not relevant, as the difference between numeric values at the endpoint were minor (see Table 3). The most frequent changes of hematological parameters (reported for > 5.0% of participants) for both groups were shifts in haematocrit from normal to low (5.5% [DRSP] and 5.4% [DSG]) or low to normal (10.8 and 11.7%) and shift in haemoglobin from normal to low (5.0 and 5.7%) or low to normal (6.2 and 6.0%). These data also reflect the similarity between the groups.

Table 3. Haematological parameters.

Parameter	Visit		DRSP 4 mg	DSG 0.075 mg	p-value
Erythrocytes (10^12/L) [R.r.: 3.6 - 5]	Baseline	Mean (SD)	4.446 (0.3107)	4.456 (0.3308)	0.6239
		Median (Min/Max)	4.430 (3.54/6.29)	4.440 (3.38/6.16)
	Endpoint	Mean (SD)	4.421 (0.3177)	4.497 (0.3382)	0.0004
		Median	4.410 (3.57/5.97)	4.470 (3.55/5.84)
	Change	Mean (SD)	−0.022 (0.2810)	0.046 (0.2773)	0.0002
		Median	−0.020 (−2.01/1.28)	0.065 (−0.86/1.44)
Haematocrit (%) [R.r.: 0.37–0.47%]	Baseline	Mean (SD)	0.396 (0.0289)	0.395 (0.0266)	0.9008
		Median	0.395 (0.30/0.52)	0.390 (0.31/0.48)
	Endpoint	Mean (SD)	0.405 (0.0303)	0.410 (0.0312)	0.0137
		Median	0.400 (0.32/0.53)	0.410 (0.31/0.49)
	Change	Mean (SD)	0.010 (0.0298)	0.015 (0.0303)	0.0043
		Median	0.010 (−0.09/0.14)	0.020 (−0.12/0.12)
Haemoglobin (mmol/L) [R.r.: 7.5 – 9.9]	Baseline	Mean (SD)	8.14 (0.612)	8.13 (0.568)	0.8920
		Median	8.10 (5.2/10.2)	8.10 (6.0/9.8)
	Endpoint	Mean (SD)	8.11 (0.574)	8.21 (0.596)	0.0068
		Median	8.10 (5.0/10.1)	8.30 (5.8/9.7)
	Change	Mean (SD)	−0.03 (0.555)	0.09 (0.515)	0.0016
		Median	0.00 (−2.4/2.7)	0.10 (−2.2/2.1)

All parameters analysed on Safetety Set population: (baseline: n = 858 for DRSP and n = 332 for DSG; endpoint: n = 838 (DRSP) and 327 (DSG)); R.r: Reference range of the respective parameter; SD: standard deviation, p-value: 2-sample t-test.

Haemostaseology

Haemostaseological parameters were evaluated in a subset of 68 participants (Table 4). All parameters were in the reference range at baseline and endpoint.

Table 4. Coagulation parametres.

Parameter	Timepoint		DRSP 4 mg	DSG 0.075 mg	p-value
Factor VII [Ref. range: 0.57–1.47%]	Baseline	Mean (SD)	1.123 (0.2486)	1.241 (0.2607)	0.0613
		Median (Min/Max)	1.120 (0.67/1.50)	1.330 (0.79/1.50)
	Endpoint	Mean (SD)	1.066 (0.2351)	1.034 (0.1964)	0.5611
		Median (Min/Max)	0.985 ((0.67/1.50)	0.980 (0.53/1.50)
	Change	Mean (SD)	−0.033 (0.2713)	−0.218 (0.2594)	0.0088
		Median (Min/Max)	−0.040 (−0.52/0.50)	−0.255 (−0.61/0.21)
Factor VIII [Ref. range: 0.64–2.1%]	Baseline	Mean (SD)	0.939 (0.2071)	1.001 (0.2075)	0.220
		Median (Min/Max)	1.000 (0.43/1.40)	1.010 (0.63/1.50)
	Endpoint	Mean (SD)	1.012 (0.2275)	1.013 (0.2035)	0.9868
		Median (Min/Max)	2.035 (0.39/1.44)	0.995 (0.66/1.50)
	Change	Mean (SD)	0.052 (0.1742)	−0.002 (0.1674)	0.2250
		Median (Min/Max)	0.030 (−0.27/0.52)	−0.015 (−0.32/0.34)
APC Resistance (Ratio) [R. r.: 2–5]	Baseline	Mean (SD)	2.711 (0.5196)	2.640 (0.4660)	0.5638
		Median (Min/Max)	2.770 (1.41/3.62)	2.700 (1.76/3.39)
	Endpoint	Mean (SD)	2.998 (0.4707)	2.989 (0.5312)	0.9458
		Median (Min/Max)	3.090 (1.99/3.98)	2.930 (2.19/4.23)
	Change	Mean (SD)	0.290 (0.5712)	0.355 (0.6461)	0.6751
		Median (Min/Max)	0.250 (−1.07/1.17)	0.410 (−0.92/2.30)
D-Dimer (ng/ml) [R.r: 0–500]	Baseline	Mean (SD)	264.9 (262.36)	201.4 (82.04)	0.2133
		Median (Min/Max)	217.0 (106/1767	196 (109/415)
	Endpoint	Mean (SD)	215.0 (110.03)	281.5 (350.04)	0.2863
		Median (Min/Max)	178.5 (74/607)	225 (68/2014)
	Change	Mean (SD)	−57.8 (306.49)	78.8 (307.99)	0.0888
		Median (Min/Max)	−25.0 (−1614/423)	18.5 (−104/1599)
Protein C Activity [R.r.: 0.7–1.5%]	Baseline	Mean (SD)	1.140 (0.2052)	1.293 (0.2447)	0.0069
		Median (Min/Max)	1.150 (0.73/1.59)	1.210 (0.91/1.78)
	Endpoint	Mean (SD)	1.108 (0.1688)	1.136 (0.2230)	0.5779
		Median (Min/Max)	1.120 (0.76/1.45)	1.135 (0.81/1.53)
	Change	Mean (SD)	−0.033 (0.2302)	−0.161 (0.1989)	0.0249
		Median (Min/Max)	−0.070 (−0.43/0.56)	−0.165 (−0.71/0.25)
AT III [R.r:0.75–1.25]	Baseline	Mean (SD)	0.946 (0.1204)	0.986 (0.1259)	0.1861
		Median (Min/Max)	0.930 (0.72/1.15)	1.000 (0.69/1.19)
	Endpoint	Mean (SD)	0.990 (0.0918)	1.033 (0.1120)	0.0948
		Median (Min/Max)	1.000 (0.76/1.14)	1.045 (0.71/1.26)
	Change	Mean (SD)	0.058 (0.1291)	0.046 (0.1034)	0.7071
		Median (Min/Max)	0.040 (0.22/0.34)	0.025 (−0.15/0.24)

Analysis was performed on a subset of participants comprising 39 (DRSP) and 29 (DSG) women at baseline and 33 (DRSP) and 28 (DSG) at endpoint. R. r: Reference range of the respective parameter. APC resistance: Resistance to activated protein C; AT III: Antithrombin III.

For all parameters (except for clotting factor VIII), changes (mean values) were more pronounced in the DSG group. The difference in change was statistically significant for clotting factor VII (Mean (SD) DRSP vs DSG: −0.033(0.2713) % vs −0.218 (0.2594) %, p = 0.0088) and protein C activity (−0.033 (0.2303) % vs −0.161 (0.2230) %, p = 0.0249). The differences between the endpoint values were minimal (mean (SD) DRSP vs DSG 1.066 (0.2351) % vs 1.034 (0.1964); p = 0.5611 for factor VII) and 1.108 (0.1688) vs 1.136 (0.2230); p = 0.5779) and therefore not considered clinically relevant.

Interestingly, a reduction in D-Dimer was observed in the DRSP dropping from 264.9 (262.36) ng/mL to 215.0 (110.03) ng/mL, whereas, in the desogestrel group, there was a rise from 201.4 (82.04) ng/mL to 281.5 (350.04) ng/mL (Mean values (SD); p = 0.0888).

Bone metabolism parameters

As markers for bone metabolism, Bone alkaline phosphatase (Bap) and Beta-CTX were determined in a subgroup of 68 participants.

The Bap level decreased over time in both groups: from 415.7 (153.51) nkat/L at baseline to 325.4 (117.74) nkat/L at endpoint for DRSP (median (SD) values are given; reference range: 193–493 nkat/L) and from 393.6 (153.06) nkat/L to 358.0 (142.65) nkat/L in the DSG group The between-group differences at baseline and at endpoint were not statistically significant (p = 0.5596 and p = 0.3250, respectively; 2-sample t-test). The mean [SD] change from baseline at endpoint was more pronounced in the Test group than in the Reference group (−79.9 [126.11] nkat/L vs. −23.9 [108.65] nkat/L). The median change was −102.0 nkat/L vs. −30.0 nkat/L, respectively. The between-group difference in mean change from baseline to endpoint was not statistically significant (p = 0.0742, 2-sample t-test).

The mean (SD) levels of Beta-CTX increased from baseline to endpoint in both groups: from 354.5 (207.51) ng/L to 366.9 (161.50) ng/L in the DRSP and from 312.9 (215.71) ng/L to 363.0 (208.11) ng/L in the DSG group (reference range: 25–573 ng/L). The mean [SD] change was 39.9 [171.08] ng/L in the Test and 47.4 [217.92] ng/L in the Reference group. The median change was 34.0 or 55.5 ng/L, respectively. The between-group difference in mean change from baseline to endpoint was not statistically significant (p = 0.8810, 2-sample t-test). The individual values of bone metabolism parameters were normal in both treatment groups.

Discussion

Findings and interpretation

The effects of hormonal contraceptives on lipid metabolism have been extensively studied for different kinds of COC. It appears to be dose-related, with a lower EE dose having a reduced metabolic impact [35], and it further reduced when less potent oestrogen derivatives are used [8–10,13–16]. Oestrogens were shown to positively affect lipid metabolism in leading to increased HDL-Cholesterol and decreased LDL-Cholesterol levels, but also lead to increased triglyceride levels. The modulating effect of progestins has also been extensively studied. Depending on their pharmacological properties, they may counteract these positive oestrogenic effects [6,22,36]: COCs with a more androgenic progestin, such as LNG, show a decrease in HDL-cholesterol and an increase in LDL-Cholesterol, while COCs with rather anti-androgenic activity, such as DRSP, were shown to have a more favourable effect on lipid profile [6,37,38].

As published recently, the impact of a DRSP/E4-containing COC on cholesterol and triglyceride levels was less pronounced compared to EE-containing COC: however, an increase in triglyceride levels (24.0% for E4/DRSP vs. 28% in EE/LNG vs. 65.5% in EE/DRSP) and in the same trial, EE/DRSP increased HDL-Cholesterol slightly, and EE/LNG decreased HDL-Cholesterol, which is in line with previous studies. E4/DRSP had no significant effect on HDL-Cholesterol, indicating that the oestrogenic effect of E4 on HDL-Cholesterol is less pronounced than the effect of EE and underlines the lower oestrogenic impact of E4 on the liver [21,38].

For progestin-only hormonal contraceptives with levonorgestrel or DSG, the effect on lipid metabolism was generally shown to be low [6, 7, 409 and this notion is confirmed by the data presented here. A slight reduction in cholesterol and triglyceride levels were observed for both the DRSP- and DSG-only contraceptive. The alteration was stronger in the DSG group compared to DRSP users, especially for triglyceride levels. This may be attributed to the more androgenic activity of DSG compared to DRSP. For long-acting progestin-only contraceptives, a similar observation was stated: implants delivering less androgenic progestins (nesterone or nomegestrolacetate) were shown to have relatively neutral effects on lipid metabolism compared to LNG implants [39]. Elevated triglyceride levels and LDL-cholesterol levels are associated with an increased risk for the development of atherosclerotic cardiovascular diseases [40]. Pathogenesis is a long-term process that is determined by different risk factors, and the epidemiological data on ATE risk is limited. Therefore, it might be worthwhile to consider the effect of hormonal contraceptives on these surrogate parameters when evaluating the individual long-term safety of HC [41].

Concerning carbohydrate metabolism, high and medium-dose COC have been associated with decreased glucose tolerance and increased insulin resistance. However, the influence of low-dose COC seems to be neglectable and is not considered essential for healthy women [15,21,42]. As demonstrated for other progestin-only contraceptives [43], no impact on carbohydrate parameters was observed in the present study for DRSP 4 mg or DSG 75 µg.

Evaluation of haemostatic parameters revealed that there was a slight decrease in erythrocyte count and haemoglobin values in the analysed DRSP subgroup compared to DSG users, and no relevant changes were observed for haematocrit. These data are applicable in the context that Folsom et al. [44] were able to demonstrate that a single high haematocrit (or haemoglobin) in a general middle-aged population sample was associated with an increased long-term VTE risk over 26 years (measured in 800 patients) compared to an average haematocrit. The highest percentile group, with haematocrits above the upper 5th percentile, had a 72% higher risk of VTE compared to the middle group. The same results were found in the description of haemoglobin. Haemoglobin levels above the upper 5th percentiles were associated with an increase in VT risk. With all other blood parameters, no relationship for the occurrence of VTE could be detected.

This, in turn, suggests that when using DRSP, the standard safety curve in terms of haematocrit and haemoglobin and thrombosis incidence shown by Folsom [44] is maintained with a substantial safety margin since DRSP 4 mg does not lead to hemoconcentration as demonstrated here.

Furthermore, the anti-mineralocorticoid potency of DRSP is approximately eight times higher than that of spironolactone (the anti-mineralocorticoid efficacy of a 4 mg DRSP tablet is roughly equivalent to that of 32 mg spironolactone) [45]. In the maintenance dose of antihypertensive therapy with spironolactone (maintenance dose: 50–100 mg to a maximum of 100–200 mg per day), dosages are applied that correspond to 6.5 times the 4 mg DRSP equivalence dose [46].

Concerning hemostaseological parameters, there was also no sign of an impact on hemostasis. D-Dimer levels were even reduced in the DRSP group compared to the DSG users. D-Dimer is a widely acknowledged marker for coagulation activation that is not only used for diagnosis of thromboembolic events [47] but also a valuable parameter in predicting the risk for recurrent VTE [48] and arterial thromboembolic events (Lowe). Therefore, the reduction in D-Dimer levels adds to the notion of cardiovascular safety of the novel DRSP 4 mg oral contraceptive.

Interestingly, it was shown that the pro-coagulatory effect of aldosterone is at least partially mediated via the mineralocorticoid receptor and can be diminished by aldosterone antagonists [49]. A strong relationship between the action of aldosterone and the induction of coagulatory parameters has been demonstrated in hypertensive patients [50]. One might speculate whether the decline in D-Dimer levels in the DRSP-group may be associated with the aldosterone-antagonistic action of DRSP. However, this is mere speculation and needs further specific studies.

No difference was observed in bone metabolism markers. For both groups, there was a decrease in bAP and an increase in CTx. Both effects were considered clinically not relevant. There is an ongoing debate about the effect of hormonal contraceptives on bone health, and up to date, no closing statement available [51,52]. For 4 mg DRSP, it was demonstrated that the serum oestrogen levels of adult users are in the range of the levels measured during the follicular phase of the normal menstrual cycle. Therefore, no negative impact on bone health is expected [52]. However, ongoing clinical trials on bone mineral density will further clarify this subject.

Strengths and weaknesses of the study

The strength of this study is the fact that the comparison was made in a high number of patients receiving different progestogens in a prospective, double-blind, double-dummy clinical trial.

The weakness of the study is that the patient collective consisted of switchers from other oral contraceptives and starters without any hormonal contraceptive use and that the starting measure point was the day of the first intake of the investigated drugs.

Relevance of the findings: implications for clinicians and policymakers

Thanks to the results, it can be concluded that the use of progestin-only pills has no clinically relevant laboratory effects showing not only their high efficacy but also the high grade of safety.

The reduction of the D Dimers in the DRSP group is a further sign of the ‘no-impact’ of the drospirenone-only pill on coagulation, reinforcing cardiovascular safety.

Conclusion

The evaluation of the laboratory data presented here supports the good cardiovascular safety profile of the novel DRSP-only oral contraceptive as reported from clinical trials: in more than 25000 application cycles, no case of VTE or ATE had been reported, although a high percentage of participants with cardiovascular risk factors had been included [29].

Disclosure statement

Santiago Palacios reports no potential conflict of interest. Pedro-Antonio Regidor and Enrico Colli are employees of Exeltis.