Introduction
Polycystic ovary syndrome (PCOS), a multifactorial disorder, involves mainly reproductive but also other endocrine systems. Based on the current Rotterdam criteria, the prevalence of PCOS varies between 5% and 10% in women of reproductive age . Metabolic issues such as inflammation, increased coagulability, visceral obesity, insulin resistance, and excess androgen are considered as key features in PCOS .
Combined oral contraceptives are hormonal medications that contain both estrogen and progestin and are used for contraceptive and noncontraceptive purposes . They are currently used by more than 100 million women worldwide. From 2015 to 2017, 12.6% of women aged 15–49 in the United States reported using oral contraception, making it the second most common method of contraception in this age range. In 2009, COCs represented 8.8% of contraceptive methods employed, reaching 15.4% in more developed countries .
Progestins are synthetic steroids possessing properties of progesterone, to which they demonstrate greater bioavailability, half-life, and activity. They exert their biological effects via binding to progesterone receptor, while they confer most of their contraceptive benefit by suppressing luteinizing hormone (LH) and ovulation. Furthermore, progestins suppress midcycle peaks of follicle-stimulating hormone (FSH) and LH, render cervical mucus hostile to spermatozoids, prevent implantation due to endometrial degeneration, slow down the movement of the oocyte, and the motility of the fallopian tubes, while they reduce mobility and fertilizing capacity of the sperm. Progestins can be distinguished, based upon their chemical origin, in those derived from testosterone and those derived from progesterone ( Table 1 ). The majority of progestins used in COCs are derived from testosterone. Depending on their chemical origin, progestins vary in their potency, androgenicity, and side effects . In contrast to the long list of progestins, only few estrogenic compounds are used in COCs, mainly ethinyl estradiol (EE) and estradiol valerate. Estrogens regulate menstrual bleeding, inhibit FSH, and prevent formation of the dominant follicle. (See Table 2 .)
| Derivatives of testosterone (19-nortestosterone: estranes and gonanes) | Derivatives of Progesterone (pregnanes) | Derivatives of spironolactone |
|---|---|---|
| Norethindrone a | Medroxyprogesterone acetate a | Drospirenone d |
| Norethindrone acetate a | Nestorone d | |
| Ethynodiol diacetate a | Cyproterone acetate a | |
| Norgestrel b | ||
| Levonorgestrel b | ||
| Norethynodrel a | ||
| Desogestrel c | ||
| Norgestimate b | ||
| Gestodene c |
a First generation progestins (low potency, well tolerated, more breakthrough bleeding with low doses of estrogen).
b Second generation progestins (greater potency, less breakthrough bleeding, more androgenic side effects).
c Third generation progestins (decreased androgenic side effects).
d Fourth generation progestins (antiandrogenic and antimineralocorticoid effects).
| Name | Properties |
|---|---|
| Monophasic pills | They deliver the same amount of estrogen and progestin each day |
| Biphasic pills | They contain estrogen and progestogens in different dose combinations |
| Triphasic pills | They contain a constant concentration of estrogen, but different concentration of progestogens during the cycle |
| Mini pill-POP | In a fixed dose |
Combined oral contraceptives exert their contraceptive effect through suppression of the hypothalamic-pituitary-ovarian (HPO) axis as well by inhibition of the zygote implantation.
The noncontraceptive effect of COCs includes regulation of the menstrual cycle and suppression of ovarian function . Thus, COCs are the most common treatment in women with PCOS. They act beneficially at multiple levels:
Pulsatility of the gonadotropin releasing hormone (GnRH) neuron: A neuroendocrine hallmark of PCOS is persistently accelerated GnRH pulsatility, which favors pituitary synthesis of LH over FSH and contributes to increased circulating LH concentrations and LH/FSH ratios, which are typical characteristics of this disorder. Kisspeptin, a hypothalamic peptide encoded by the KiSS1 gene, is a neuromodulator and a potent stimulator of the HPO axis. Secreted upstream of GnRH, it signals directly to GnRH neurons via kisspeptin receptor to release GnRH into the portal circulation resulting to pulsatile secretion of LH and FSH from the anterior pituitary . Estrogen and progesterone, both components of COCs, modulate kisspeptin activity through sex steroid receptors . An undefinable period of treatment with COCs might lead to normalization of GnRH pulsatility.
Inhibition of endocrine secretion from the ovaries: The indirect, via suppression of the HPO axis, inhibition of ovarian endocrine function by COCs results in suppression of ovarian androgens secretion and subsequently in reduction of symptoms of hyperandrogenism. The decrease of androgens contributes, in turn, to the increase of unbound progesterone receptors in the hypothalamus, leading, thus, to normalization of GnRH pulsatility.
Other effects: Estrogens in COCs lead to increased concentrations of sex-hormone binding globulin (SHBG), which binds, among other peripherally circulating hormones, androgens. Thus, the amount of circulating free androgens decreases. Progestogens in COCs compete with androgens for the binding of the androgen receptors and inhibit 5a reductase, which catalyzes the reduction of testosterone to its active metabolite, dihydrotestosterone . Specific progestogens possess particular antiandrogenic properties by binding and blocking androgen receptors. In this way, COCs improve hair loss and acne . Also, progestogens stabilize and protect the endometrium from the constant exposure to estrogens as it is the case in PCOS.
Pharmacology of COCs
Combined oral contraceptives are administrated per os. Their side effects are those of estrogen and progestins. Enovid®, the first COC, contained mestranol and norethisterone, and it was approved by the Food and Drug Administration in May 1960. The addition of an ethinyl group to the estradiol molecule in 1938 created EE, the synthetic estrogen compound most used in all COCs. Over the years, COCs have evolved through gradual lowering of EE content, introduction of 17β estradiol, and many different progestins.
There are three molecularly distinct types of progestins according to their active chemical structure and their parent compound: estranes, gonanes, and pregnanes ( Table 1 ). The former two, derived from testosterone, were created with the addition of an ethinyl group to the molecule of testosterone to create ethisterone. The subsequent removal of a carbon atom from the 19th carbon of ethisterone converted it from an androgen to a progestogen (19-nortestosterone derivatives). They include norethindrone, norethindrone acetate, ethynodiol diacetate, norethynodrel (estranes) and levonorgestrel (LNG), desogestrel (DSG), norgestrel, norgestimate (NGM), and gestodene (GSD) (gonanes) ( Table 1 ). Gonanes and estranes differ in their half-life as well as with respect to their estrogenic and antiestrogenic effects. Pregnanes derive from progesterone. They include cyproterone acetate (CPA), nestorone, and medroxyprogesterone acetate. Drospirenone (DRSP) is a spironolactone analog with antimineralocorticoid and antiandrogenic activity . The COCs are also classified according to the time they were introduced into the market in four generations ( Table 1 ). Combined oral contraceptives contain EE or estradiol valerate as an estrogen. Oral contraceptives have been related to deterioration of lipid profile although without significant changes in other metabolic outcomes, including body mass index (BMI), fasting blood glucose, fasting insulin, homeostatic model for measuring insulin resistance (HOMA-IR), and blood pressure values. Combined oral contraceptives with an antiandrogenic progestin (CPA, DRSP) are employed in the treatment of PCOS and for decades have been the most common therapeutic tool for PCOS ( Table 1 ) . Cyproterone acetate (CPA) demonstrates significant progesterone activity regarding inhibition of gonadotropin secretion. Drospirenone in therapeutic doses has antiandrogenic and mild antimineralocorticoid properties. It has no estrogenic, glucocorticoid, and antiglucocorticoid action, while it exhibits a pharmacological profile quite similar to progesterone .
Side effects and contraindications of therapy with COCs
Side effects
Nausea, breast tenderness, and headaches are usually minor and infrequent complaints (< 10% of women) after initiation of COCs and are less common with current formulations . Side effects include unscheduled bleeding, mood changes, and increased risk for cervical cancer. Usually, these symptoms resolve within the first months of use; however, they can sometimes be bothersome enough to cause discontinuation shortly after COC initiation . Unscheduled bleeding commonly occurs with COC use, but it improves over the first 3–4 months of use . Amenorrhea occurs with COCs with low EE concentrations (relative to the much greater progestin doses), which are inadequate to stimulate endometrial growth, resulting in the absence of withdrawal bleeding. Amenorrhea may also occur rarely after discontinuation of COCs but is not a consequence of COC use itself and does not indicate that a patient is not ovulatory . Regarding weight gain, the analysis of four randomized control trials (RCT) revealed no difference in body weight gain after 6 months of COCs use in non-PCOS women . An observational study following COCs users for more than two decades also did not demonstrate significant weight gain . Evidence suggests that COCs efficacy is similar for normal-weight and overweight/obese women. Combined oral contraceptives can impact negatively lipid and carbohydrate metabolism but usually not in a clinically meaningful manner . However, for patients with PCOS, these changes should be given special attention (see below). Further, there is no evidence that the use of COCs influences the risk of developing diabetes mellitus . Regarding cervical cancer, a collaborative analysis of data from 24 worldwide observational studies, including large UK cohort studies, suggested that current use of COCs for more than 5 years approximately doubled the risk of invasive cervical cancer [RR 1.90; 95% confidence intervals (CI) 1.69–2.13] compared with never-use of COCs . The risk declined after stopping COCs, becoming the same as that for never-users about 10 years after cessation. Inversely, COCs use is associated with a decreased risk for ovarian (incidence rate ratio 0.67; 99% CI 0.50–0.89; with protection lasting for over 30 years) and endometrial cancer (incidence rate ratio 0.66; 99% CI 0.48–0.89) .
Contraindications
The WHO Medical Eligibility Criteria for Contraceptive Use lists contraindications to the use of COCs , including hypertension, smoking over age of 34, and migraine with aura, among others. Combined oral contraceptives are advised not to be used or used with caution in the presence of the following conditions:
- •
thrombophlebitis and thromboembolic disorders or history of venous thrombosis (deep venous thrombosis, pulmonary embolism);
- •
presence or history of arterial thrombosis (myocardial infarction, cerebrovascular accident) or prodromal conditions (e.g., transient ischemic attack, angina pectoris);
- •
history of migraine with focal neurological symptoms;
- •
known predisposition for venous or arterial thrombosis, such as activated protein C (APC) resistance, antithrombin-III deficiency, protein C deficiency, protein S deficiency, hyperhomocysteinaemia, and antiphospholipid antibodies;
- •
valvular heart disease with complications;
- •
severe hypertension (persistent systolic values of > 160 mmHg or persistent diastolic values of > 100 mmHg);
- •
diabetes mellitus with vascular involvement;
- •
presence or history of severe hepatic disease as long as liver function values have not returned to normal;
- •
presence or history of liver tumors (benign or malignant);
- •
known or suspected malignant conditions of the genital organs or the breasts;
- •
endometrial hyperplasia;
- •
undiagnosed vaginal bleeding;
- •
cholestatic jaundice of pregnancy or jaundice with prior pill use; and
- •
known or suspected pregnancy.
Why treat PCOS with COCs?
Combined oral contraceptives have been employed for many years as a first-line treatment in women with PCOS to reduce manifestations of hyperandrogenism and regulate the menstrual cycle . Specifically, COCs contribute to the following:
Amelioration of hyperandrogenism (hirsutism, acne, scalp hair loss) . Combined oral contraceptives inhibit ovulation through suppression of hypothalamic GnRH and suppression of pituitary gonadotropin secretion. Suppression of LH corrects one of the etiologic features of PCOS physiopathology, while secretion of FSH is also decreased although to a lesser extent . In studies of COCs-induced gonadotropin suppression in women with PCOS, serum FSH concentrations fell more rapidly than did serum LH concentrations, but the fall in serum LH concentrations was greater (70% vs. 50%) . As a result, secretion of ovarian testosterone and Δ 4 androstenedione decreased substantially. Furthermore, estrogens and progestins in COCs lead to a dose-dependent increase in SHBG. The latter binds circulating testosterone and estradiol with high affinity. In women, only 1% to 1.5% of serum testosterone represents the biologically active free form . Lower serum SHBG concentrations than normal are often documented in PCOS women, as a consequence of increased androgen and insulin concentrations, both of which decrease its hepatic production . Both inhibition of LH secretion and increase in SHBG production contribute to decrease of serum free testosterone concentrations by approximately 50% in PCOS women treated with COCs . Serum total testosterone concentrations also fall, but not as much as free testosterone, because of the increase in serum SHBG concentrations. These changes may occur over several weeks but can become apparent later, in case of women with marked insulin resistance in whom production of both ovarian androgen and hepatic SHBG presumably depends more on insulin than LH or estrogens . In addition, COCs can decrease adrenal androgen secretion via suppression of 5α reductase activity and subsequent decrease of the conversion of testosterone to dihydrotestosterone as well as inhibition of dihydrotestosterone binding to androgen receptors. However, there is not enough evidence to support the contribution of these effects to the reduction of hyperandrogenism in women with PCOS.
Prevention of endometrial hyperplasia and carcinoma, which may occur as a result of chronic hyperestrogenemia.
Contraception for women not seeking pregnancy . Women with PCOS ovulate irregularly, and unwanted pregnancy may occur.
When treat PCOS with COCs?
Adolescence. Combined oral contraceptives usually are the first-line treatment for adolescents with PCOS and abnormal menstrual bleeding or cutaneous signs of excess androgen . This is the suggestion of both the International Society of Endocrinology and the Endocrine Society guidelines. The diagnosis of PCOS during adolescence (the period of life between 10 and 19 years of age according to WHO) is both controversial and challenging due to the overlap of normal pubertal physiological changes (irregular menstrual cycles, acne, and polycystic ovarian morphology) with adult PCOS diagnostic criteria. As a general rule, COCs should be continued until the patient is gynecologically mature (5 years postmenarchal) or has lost a substantial amount of excess weight. At this time, a trial-off of therapy is reasonable to document the persistence of the syndrome . Adolescent girls at risk of PCOS who do not meet all diagnostic criteria might also be treated with COCs, but they should be reassessed at 8 years postmenarche to check if PCOS features persist. These girls should be transitioned appropriately from pediatric to adult care . Besides the known limitations for the use of COCs (i.e., increased risk of VTE, increased risk of thrombotic stroke and myocardial infarction, promotion of salt and water retention, desire for pregnancy), one must consider the risk of growth inhibition, by the COC estrogen, in perimenarchal girls with short stature for whom growth potential is important . Concerns have been raised that the incompletely mature adolescent neuroendocrine system may present increased risk for postpill amenorrhea . However, this hypothetical concern is based on observations in patients treated with high-dose estrogens during adolescence. Most postpill amenorrhea is thought to reflect the presence of an undiagnosed preexisting condition, such as PCOS or hyperprolactinemia.
Adulthood . In adult PCOS women, COCs are the mainstay of pharmacologic therapy for managing hyperandrogenism and menstrual dysfunction and for providing contraception. Women may use COCs until the age of 50 years, provided there are no contraindications. After the age of 50 years, caution is warranted because usually the risks outweigh benefits and should be decided on an individual basis. Use of COCs carry an increased risk for VTE per se, particularly in obese women, while there is additional risk for VTE after 40 years of age.
Where treat PCOS with COCs
According to the severity of hirsutism
For the evaluation of hirsutism, which is excessive terminal hair that appears in a male pattern in women, the gold standard has been the modified Ferriman-Gallwey score (mFG) .
Recommended treatment with COCs according to the type of progestin: Most progestins are derived from 19-nortestosterone and exhibit varying degrees of androgenicity . Norethindrone exhibits medium androgenicity while norgestrel and LNG exhibits relatively high androgenicity. Cyproterone acetate and DRSP are progestogens contained in widely used COCs for the treatment of PCOS. They are structurally unrelated to testosterone and function as weak androgen receptor antagonists. Drospirenone is structurally related to spironolactone, which exhibits weak antiandrogenic activity. In bioassays measuring antiandrogenic activity, 3 mg DRSP (the dose used in COCs) is equivalent to 9–10 mg spironolactone. For comparison, 100 to 200 mg spironolactone is the therapeutic dose for hirsutism. Also, 2 mg CPA (the dose used in COCs) is equivalent to 50 mg spironolactone . A 12-month trial comparing COCs containing either 3 mg DRSP or 2 mg CPA showed similar reductions in hirsutism scores, suggesting that their antiandrogenic efficacy is substantially related to ovarian suppression .
In a 2018 Endocrine’s society systematic review, COCs containing either antiandrogenic progestins (CPA and DRSP) or the relatively androgenic progestin LNG were compared to all other COCs . Only four of the included trials presented data appropriate for meta-analysis. The LNG containing COCs showed a similar effect on hirsutism scores compared with all other COCs. The COCs containing the antiandrogenic progestins (one trial using CPA and one trial using DRSP) were associated with slightly lower mFG scores compared to those of other COCs, with a weighted mean difference not clinically important. A potential benefit of COCs containing LNG is their lower VTE risk, whereas adverse effects of LNG on metabolic biomarkers are observed . Ιn women demonstrating hirsutism and are at higher risk for VTE, the Androgen Excess and PCOS Society suggests an initial therapy with a COC containing the lowest effective dose of EE (usually 20 μg) and a progestin low-risk for VTE . Regarding suppression of the HPO axis by the estrogens contained in COCs and the subsequent decrease of circulating ovarian androgens, in a meta-analysis of 42 studies, suppression of serum total and free testosterone concentrations was comparable among COCs containing 20, 30, or 35 μg EE . Limited data suggest that COCs containing DRSP with either 20 or 30 μg EE have a similar effect on mFG scores .
If the clinical response to COCs containing a neutral progestin is unsatisfactory, changing the COC formulation to include an antiandrogenic progestin may be useful. Of note, these low-dose third-generation COC formulations are not associated with the unfavorable metabolic profile of older formulations and may even have beneficial effects on the lipid profile even among obese insulin-resistant patients with PCOS . However, a mild increase in blood pressure might occur with some of these newer COCs . Especially among smokers, the third-generation COCs may have deleterious effects on coagulation and may increase the risk of nonfatal VTE compared with second-generation COCs with LNG .
According to the severity of obesity
In obese perimenopausal women, COCs should be used with caution, as the risk of VTE increases with BMI ≥ 30 kg/m 2 ; risk appears to be twice as high as for nonobese women. Of note, most studies include relatively few women in the greatest weight or BMI categories. Obese women who are not perimenopausal can use COCs, as the benefits are generally believed to outweigh risks.
Eleven studies have examined the 3- to 6-month impact of COCs on body weight in PCOS patients . Two studies employing COCs with EE and CPA demonstrated a nonsignificant body weight increase suggesting that this COC does not influence body weight in PCOS users. The use of COC with 30 μg EE and 3 mg DRSP resulted in a nonsignificant body weight decrease . Three studies employing a COC with 30 μg EE and 2 mg chlormadinone acetate (CMA) in PCOS patients demonstrated a nonsignificant increase in body weight . PCOS patients who used a COC with 30 μg EE and 150 μg DSG had a nonsignificant increase in body weight . In summary, most of the COCs studied did not promote significant weight gain in PCOS patients.
Thirteen studies evaluated the impact of COCs with 35 μg EE and 2 mg CPA on BMI of PCOS patients. Four and four studies demonstrated a significant and a nonsignificant decrease in BMI, respectively. Six studies reported a nonsignificant increase in BMI. Collectively, these studies demonstrated a slight nonsignificant decrease of BMI in PCOS patients who use EE combined with CPA . BMI changes in PCOS users of COCs with 30 μg EE and 3 mg DRSP were reported in 17 clinical studies. Two and seven of them reported a significant and a nonsignificant BMI decrease, respectively. Six studies reported a nonsignificant increase in BMI. Collectively, these studies demonstrated a nonsignificant increase in BMI with the use of COC with DRSP . Four studies and one study using a COC with EE and CMA in PCOS patients reported a nonsignificant BMI increase and decrease, respectively. Taken together, this COC does not modify BMI of PCOS patients . In seven studies, the use of a COC with EE and DSG was shown to decrease BMI insignificantly. In eight and another one study employing a COC with EE combined with DSG, a nonsignificant and a significant increase, respectively, in the BMIs of PCOS patients was demonstrated . One study employing a COC with EE and GSD demonstrated a nonsignificant increase in BMI . Recently, one study employing a COC with EE and NGM demonstrated a nonsignificant decrease in BMI. In conclusion, current studies do not endorse the idea that COCs increase BMI in PCOS patients.
According to the cardiovascular background of PCOS patients
Patients with PCOS are suspect to increased risk for VTE and cardiovascular disease (CVD). Then, it is imperative to consider the possible existence of additional cardiovascular risk factors for these patients, including the use of COCs . Use of COCs has also been associated with a twofold increased risk of fatal and nonfatal cardiovascular events, which are perhaps related to endothelial dysfunction due to altered levels of nitric oxide, homocysteine, and angiotensin . It should be emphasized that the association between the use of COCs and CVD has been demonstrated in non-PCOS subjects, mainly among women who are smokers and/or are over 35 years old and/or are users of third- or fourth-generation COCs, which contain less androgenic progestins . Three studies have examined the risk of VTE in users of COCs with PCOS. One study using a combination of 35 μg EE and 2 mg CPA detected an odds ratio of 2.2 (95% CI, 1.35–3.58) and 7.44 (95% CI, 3.67–15.08) for the risk of VTE compared with users of other COCs and to nonusers, respectively . Interestingly, in a cross-sectional study of women 18–45 years old who used any kind of COC, a protective association (OR = 0.8; 95% CI, 0.73–0.98) over VTE in PCOS users was reported . On the other hand, a Canadian study that included women who were 18–46 years old and used different criteria for PCOS diagnosis and different COCs containing < 35 μg EE combined with second- and third-generation progestins found a twofold increased risk for VTE in PCOS users (RR = 2.14, 95% CI, 1.41–3.24), a risk even greater in women with hyperandrogenism (HR = 2.49; 95% CI, 1.35–4.59) . These studies have limitations because they did not use the same criteria for PCOS diagnosis, and they compared subjects using a different combination of COCs and did not control for confounding variables such as PCOS phenotype, diet, exercise, or body mass index.
The role of blood pressure. Regarding the effects of COCs on systolic blood pressure (SBP) in PCOS patients, 14 studies examined the role of almost all COCs formulations . Four studies using a COC with 35 μg EE and 2 mg CPA demonstrated a nonsignificant increase in SBP. Seven studies using a COC with 30 μg EE and 3 mg DRSP demonstrated a nonsignificant decrease in SBP, with the exception of the Battaglia et al. study which demonstrated a significant rise in SBP. Among the three studies employing a COC with 30 μg EE and 2 mg CMA in PCOS patients, two and one showed a nonsignificant decrease and increase in SBP, respectively. A nonsignificant increase in SBP was also demonstrated in two studies using a COC with 30 μg EE and 75–150 μg DSG. Two studies using a COC containing NGM demonstrated no change in SBP. In summary, the majority of available data indicate nonsignificant changes in baseline SBP with the use of COCs in PCOS patients, independently of the formulation.
Regarding the effects of COCs on diastolic blood pressure (DBP) in PCOS patients, 12 studies examined the role of almost all COCs formulations . Five studies using a COC with 35 μg EE and 2 mg CPA demonstrated a nonsignificant increase in DBP. Choosing this formulation for PCOS patients with arterial hypertension might not be recommended. Among the seven studies employing a COC with 30 μg EE and 3 mg DRSP, five and two demonstrated a nonsignificant increase and decrease, respectively, in DBP. Three studies using a COC with 30 μg EE and 2 mg CMA demonstrated a nonsignificant decrease in DBP. Two studies using a COC with 30 μg EE and 75–150 μg DSG demonstrated a nonsignificant increase in DBP. Collectively, it could be concluded that COCs, independently of the type of combination, do not significantly influence DBP in PCOS patients. However, COCs with CMA might be preferred in PCOS patients with arterial hypertension.
The role of lipids. The impact of COCs on lipid metabolism in PCOS patients has been investigated in several studies.
Total cholesterol. The influence of COCs containing 35 μg EE with 2 mg CPA on total cholesterol (TC) concentrations in PCOS patients has been investigated in a number of studies. Five and another five studies reported nonsignificant and significant increases, respectively, of TC concentrations. In addition, one and another one study reported a nonsignificant and a significant decrease, respectively, of TC concentrations. The analysis of these 12 studies as a whole demonstrated that TC concentrations increased significantly from 173.7 mg/dL to 189.1 mg/dL. Thus, it appears that the use of COC containing CPA in PCOS patients causes a significant increase of TC concentrations . Thus, its use in dyslipidemic PCOS patients should not be recommended. The role of COCs with DRSP on the TC concentrations in PCOS patients has been examined in nine studies. Two and six studies reported nonsignificant and significant increase of TC, respectively, while one study reported a significant decrease. Taken together, all nine studies presented a nonsignificant increase of TC concentrations . In one study, a COC with EE and CMA showed a significant increase in TC concentrations from 161.1 mg/dL to 193.2 . In PCOS patients, COCs with 30 μg EE and either 75–150 μg DSG or 75 μg GSD or 180–250 μg NGM or 150 μg LNG led to a nonsignificant increase of TC concentrations . Collectively, the results of the existing studies indicate that certain COCs increase TC concentrations in PCOS patients. This fact must be considered for proper COC prescription in dyslipemic conditions.
HDL cholesterol (HDL-C). Taking together all of 10 studies using COCs containing 35 μg EE and 2 mg CPA , it appears that they increased significantly HDL-C concentrations from 54.3 mg/dL to 61.5 mg/dL. In three of these studies, HDL-C concentrations presented a nonsignificant increase while in one of them HDL-C concentrations did not change. In PCOS patients, in eight studies, a COC with 30 μg EE and 3 mg DRSP increased significantly HDL-C concentrations from 59.5 to 73.3 mg/dL , while in two studies, it increased HDL-C concentrations nonsignificantly. Three studies using COCs with 2 mg CMA and three studies using COCs with DSG in PCOS patients reported a nonsignificant increase and decrease of HDL-C concentrations, respectively . Also, a COC with 30 μg EE and 75 μg GSD presented a nonsignificant decrease in HDL-C concentrations . Collectively, these results indicate that the increase in HDL-C concentrations with most COC preparations represents a favorable clinical consequence.
LDL cholesterol. The concentrations of LDL-C in PCOS patients treated with COCs withed CPA was examined in 11 clinical studies . Three and five of them reported nonsignificant decrease and increase, respectively, in LDL-C concentrations. Two studies and one study reported a significant increase and decrease, respectively, in LDL-C concentrations. As a whole, these studies demonstrated that following administration of COCs with CPA, LDL-C concentrations increased nonsignificantly. Results regarding LDL-C concentrations in PCOS patients treated by COCs with DRSP are inconsistent ranging from nonsignificant to significant decrease and from nonsignificant to significant increase. As a whole, these studies regarding LDL-C concentrations, following administration of COCs with DRSP, were inconsistent . Furthermore, studies have shown that following administration of COCs with CMA or DSG, LDL-C concentrations decreased and increased nonsignificantly, respectively . One study reported a nonsignificant increase in LDL-C concentrations, in PCOS patients treated with COCs with GSD . COCs with EE 30 μg and either NGM 180–250 μg or 150 μg LNG increased significantly LDL-C concentrations from 100.2 to 107.6 mg/dL and from 87.2 to 90.6 mg/dL, respectively . From these studies, it appears that the effect of COCs on the LDL-C levels is inconsistent, especially with those containing CPA or DRSP, while those containing either LNG or NGM might have a more negative effect.
Triglycerides. The impact of COCs containing CPA on triglycerides (TG) concentrations is reported in 11 clinical studies. In seven of them, a significant increase in TG concentrations was reported while in two other studies, a nonsignificant one. While in one study TG concentrations did not change following administration of this COC, in one study a significant decrease in TG concentrations was reported. The analysis of all 11 studies with COCs containing CPA revealed that TG concentrations increased significantly from 81.4 to 104 mg/dL . Concentrations of TG in PCOS patients treated with COCs with DRSP were examined in seven clinical studies. Six and one of them reported significant and nonsignificant increase in TG concentrations, respectively. As a whole, these studies demonstrated that TG concentrations increased significantly from 95 to 116.8 mg/dL . Following administration of COCs containing either CMA or DSG or GSD or LNG, TG concentrations increased non-significantly , while following the administration of COCs containing NGM, TG concentrations increased significantly from 104.7 to 134.3 mg/dL . Collectively, it appears that TG concentrations are mostly increased by the use of COCs. Therefore, the presence of hypertriglyceridemia should be considered before any COC prescription. Combinations containing CMA or DSG or GSD or LNG seem to present a lower impact on TG concentrations in PCOS women.
The role of carbohydrate metabolism
Fasting glucose. Regarding the impact of COCs on the carbohydrate metabolism in PCOS patients, 12 studies published between 1990 and 2016 examined the role of COC with 35 μg EE and 2 mg CPA on fasting glucose. Seven of them demonstrated a nonsignificant increase in fasting glucose. Three, one, and one of them demonstrated, respectively, no change, nonsignificant decrease, and significant decrease of 8.0%. As a whole, these studies demonstrated that following administration of COC with CPA, fasting glucose concentrations decreased nonsignificantly . One study demonstrated a nonsignificant decrease of the glucose area under the curve (AUC) in a 3 h oral glucose tolerance test (OGTT) following administration of a COC with CPA in PCOS patients . Ten studies examined the role of COC with 35 μg EE and 3 mg DRSP on fasting glucose. Four and three of them demonstrated a nonsignificant increase and decrease, respectively, in fasting glucose. Two and one of them demonstrated a significant decrease and no change, respectively, in fasting glucose. As a whole, these studies demonstrated that following administration of COC with DRSP, fasting glucose concentrations increased nonsignificantly . Two and one studies examined the role of a COC with DRSP on glucose AUC following OGTT in PCOS patients, and they demonstrated nonsignificant and significant increase, respectively, of glucose AUC . Thus, it appears that the use of COC containing DRSP in PCOS patients is harmless on glucose metabolism. Two studies demonstrated a significant decrease in fasting glucose in PCOS patients treated by COC with 35 μg EE and 2 mg CMA . Five studies demonstrated a nonsignificant increase in fasting glucose in PCOS patients treated by COC with 35 μg EE and 150 mg DSG . Three studies demonstrated a nonsignificant decrease in fasting glucose in PCOS patients treated by COC with 33 μg EE and 180–250 μg NGM . Thus, it appears that the use of COC with either CMA or DSG or NGM in PCOS patients does not significantly modify fasting glucose concentrations.
Fasting insulin. The impact of COCs with 35 μg EE and 2 mg CPA on fasting insulin is reported in 11 clinical studies. In four and two of them, a nonsignificant and a significant decrease in fasting insulin was reported, respectively. In one and four others of them, no change and a nonsignificant increase in fasting insulin was reported, respectively. As a whole, these studies demonstrated that following administration of COCs with CPA, fasting insulin concentrations decreased nonsignificantly . Two studies demonstrated significant decrease in the insulin AUC following OGTT in PCOS users of a COC with CPA . Thus, it appears that the use of COC with CPA in PCOS patients does not worsen insulin metabolism. The impact of COCs with 30 μg EE and 3 mg DRSP on fasting insulin is reported in eight clinical studies. In six of them, a nonsignificant increase in fasting insulin was reported, while in another one, an astonishing 241% increase was reported, although the latter had a very small sample size and a very low power. While one study reported a nonsignificant decrease in fasting insulin, significant decreases were also demonstrated in other studies. As a whole, these studies demonstrated that following administration of COCs with DRSP, fasting insulin concentrations decreased nonsignificantly . In one study and in another one, a very small clinically nonsignificant and a clinically significant 20.3% decrease of the insulin AUC following OGTT was reported, respectively, while in another study, a nonsignificant increase of the insulin AUC was reported. Thus, it appears that the use of COCs with DRSP in PCOS patients do not worsen insulin metabolism. Following administration of COCs with 30 μg EE and 2 mg CMA, fasting insulin concentrations increased nonsignificantly . As a whole, following administration of COCs with 30 μg EE and 75–150 DSG, fasting insulin concentrations increased nonsignificantly . The administration of COCs with 30 μg EE and 180–250 μg NGM fasting insulin concentrations increased nonsignificantly in two studies . Conclusions drawn from studies reporting the impact of COCs on fasting insulin concentrations indicate that COCs with either CMA, or DSG or NGM promote a small nonsignificant increase in fasting insulin concentrations in PCOS patients.
Glycated hemoglobin (HbA1c). Glycated hemoglobin may be a reliable marker of overall glycemia status in PCOS individuals . Its values are increased in PCOS patients as compared with non-PCOS controls . Changes in the HbA1c of PCOS patients using COCs were reported in a few studies. The impact of COCs with 35 μg EE and 2 mg CPA on HbA1c is reported in four clinical studies, which demonstrated collectively a nonsignificant increase in HbA1c . Similarly, the administration of a COC with EE 30 μg and DRSP 3 mg did not change HbA1c values in 6 months.
Considering the different phenotypes of PCOS, PCOS patients may present with a slight increase in blood pressure, be overweight or obese, and present with dysglycemia and dyslipidemia. The use of different COC preparations may result in some harmful effects. Thus, the phenotypes of PCOS should be taken into consideration, and the choice of a COC should be based on the estrogen type, dose, and the possible androgenic properties of the progestin.
Side effects of treating PCOS with COCs
The use of COCs has been associated with increased thromboembolism and hypertension rate, changes in lipid and carbohydrate metabolism, depending on the estrogen dose and the androgenicity of the progestin. However, there is no evidence that women with PCOS treated with COCs are at greater risk for either metabolic adverse effects or severe cardiovascular complications. Collectively, the available studies, either randomized controlled trials or prospective observational ones, suggest that the use of different COCs preparations do not cause particular harm in PCOS users, at least when used for short periods of time (3–6 months). The relatively short protocol period of the published studies is a major disadvantage in drawing significant evidence-based conclusions regarding treatment of PCOS patients with COCs. Because of remarkable decrease in androgen and increase in SHBG concentrations, patients presenting with the hyperandrogenemic phenotype may demonstrate greater benefits compared to the possible metabolic side effects of these formulations. Antiandrogenic progestins are preferred in patients with this phenotype.
Two studies based upon databases of health insurance claims reported that PCOS itself appeared to be an additional risk factor for VTE. However, both had important methodological limitations, and it cannot be determined whether the excess risk of VTE in this population was due to PCOS or obesity and other comorbidities. Neither study directly addressed the question of whether the disproportionate use of COCs might increase the VTE risk in women with PCOS. For most women with PCOS, the benefits of COCs still outweigh the potential risks of VTE. Furthermore, although PCOS patients with hypertension, or central obesity, can use any COC preparation, solid evidence for adverse effects is missing. Scarce data suggest that preparations containing antiandrogenic/antimineralocorticoid progestin are preferable in these cases.
Estrogens are known to decrease LDL and increase TC, HDL, TG, and SHBG. Total cholesterol, TG, and HDL-C are consistently increased by COCs, but the clinical implications of these changes still need more long-term investigation. Thus, lipid baseline concentrations should be measured to help choosing a COC preparation for dyslipidemic patients.
Regarding changes in carbohydrate metabolism indices in PCOS patients using COCs evidence-based data indicate that COCs do not lead to significant deterioration of their baseline values. In healthy women, the use of COCs decreases insulin sensitivity, but in general, this decrease is not clinically significant. Thus, it has been extrapolated that the use of COCs would worsen insulin sensitivity as well in women with PCOS. However, data are conflicting, with studies showing improvement, deterioration, or no change in insulin sensitivity. Of note, COCs with DRSP and DSG may decrease HOMA-IR and improve tissue insulin sensitivity.
Conclusion
Combined oral contraceptives are the most common treatment in women with PCOS. Besides their contraceptive action, they act therapeutically at multiple levels, mainly in the pulsatility of the GnRH neuron and in the inhibition of hormonal secretion from the ovaries. Combined oral contraceptives contain EE or estradiol valerate as an estrogen and different types of progestins. Initiation of treatment with COCs is associated with side effects such as nausea, breast tenderness, and headaches, which are usually minor and infrequent complaints (< 10% of women), being however less common with current formulations . Unscheduled bleeding, when manifesting as side effect, resolves with time . Combined oral contraceptives have not been found to associate with weight gain in large RCTs . However, certain COCs can impact negatively on lipid and carbohydrate metabolism . As far as the lipid profile of COCs in these patients, the existing studies indicate that most preparations increase TG, while it does not seem that a preferred formulation exists. Additionally, the use of most COC preparations has been associated with an increase in HDL-C and TC, but the effect on LDL-C concentrations is inconsistent. Regarding the effect on the carbohydrate metabolism, it appears that the use of COCs does not significantly modify fasting glucose concentrations, HbA1c or fasting insulin concentrations. Use of COCs is associated with decreased risk for ovarian and endometrial cancer and increased relative risk for invasive cervical cancer depending on the years of use . Breast cancer and history of VTE and CVD represent absolute counterindication for their administration . Of note, a mild increase in blood pressure might occur with some of these newer COC formulations . Especially among smokers, third-generation COCs may have deleterious effects on coagulation and may increase the risk of nonfatal VTE compared with second-generation COCs containing LNG . Οvarian androgen suppression by COCs appears to be similar with the different doses of EE contained in COCs.
Combined oral contraceptives have been employed for many decades as a first-line treatment in women with PCOS counteracting the manifestations of hyperandrogenism (hirsutism, acne, scalp hair loss), contributing to prevention of endometrial hyperplasia and carcinoma, which may occur as a result of chronic hyperestrogenemia, and regulating menstrual cycle. In adolescence, COCs, which contain estrogen and progestin, usually are the first-line treatment for PCOS patients with abnormal menstrual bleeding or cutaneous signs of hyperandrogenism . As a general rule, COCs should be continued until the patient is gynecologically mature (5 years postmenarchal), given the consideration of the risk of growth inhibition, by the COC estrogen, in perimenarchal girls with short stature for whom growth potential is important . In adult PCOS women, COCs are the mainstay of pharmacologic therapy for managing hyperandrogenism and menstrual dysfunction. They can be used until the age of 50 years, after which caution is mandatory because risks might outweigh benefits. Thus, tailoring of treatment with COCs is advised.
Most available progestins have varying degrees of progestogenic and androgenic activity. However, in COCs, when combined with estrogen, the overall effect is antiandrogenic. In general, third- and fourth-generation progestins are less androgenic and have fewer side effects than earlier generation progestins. However, some of these later generation progestins (DSG and NGM) have also been associated with a slightly greater risk of VTE. Progestins characterized for their antiandrogenic properties (CPA and DRSP) represent a special category of COCs employed in the treatment of PCOS. Interestingly, the effect of either the more androgenic (e.g., LNG) or the antiandrogenic progestins (CPA and DRSP) in COCs is similar on hirsutism score. A potential benefit of COCs containing LNG is their lower VTE risk . Although no clear data suggest that the metabolic effects of the use of LNG in COCs are associated with adverse clinical outcomes, one tends to balance therapeutic options including LNG in women with PCOS, a population at metabolic risk. Of note, the antiandrogenic progestins (CPA and DRSP) are not associated with the unfavorable metabolic profile of older formulations . A mild increase in blood pressure might occur with some of these newer COCs . Regarding the additive effect on the risk for VTE posed by PCOS per se, no studies have directly addressed the question of whether the use of COCs might increase VTE risk disproportionately in these patients. Furthermore, although PCOS patients with hypertension, or central obesity, can use any COC preparation, evidence on their role in these particular patients is missing. Scarce data suggest that preparations containing antiandrogenic/antimineralocorticoid progestin could be preferable.
Therefore, the choice of a particular COC for the treatment of PCOS must carefully balance the greater efficacy of third-generation COC against the safer coagulation profile of second-generation COC, especially in adolescents, hypertensive women, obese women, and smokers. Regarding the side effects of the different COC formulations in PCOS patients, it appears that they do not promote significant weight gain or have a significant influence in blood pressure. It appears that preparations containing antiandrogenic/antimineralocorticoid progestin could be favored. Further, the phenotypes of women with PCOS should be taken into consideration, and the decision to choose a specialized COC formulation should be based on the estrogen type, dose, and the possible androgenic properties of the progestin. Potential risks or disadvantages of COC use in PCOS women should be evaluated based on published evidence.
The treatment option that COCs offer in PCOS should be tailored according to needs and preferences of each patient. Goals of treatment are to improve quality of life and long-term health outcomes.
References
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
