This chapter should be cited as follows:
Update due

Pharmacology of Contraceptive Steroids

Authors

INTRODUCTION

Contraceptive steroids represent a major method of birth control in the United States and in many other countries. Because of their effectiveness and good patient acceptability, the oral progestin/estrogen combinations are the most widely used steroidal formulations for contraception. A constant ratio of progestin/estrogen is most common, although preparations with increasing proportions of the progestin component have gained some advocates. Sequential preparations in which unopposed estrogen is given in the first part of the cycle have been discontinued in many parts of the world. The use of the 'minipill' of continuous progestin is limited because it is associated with a high incidence of intermenstrual bleeding and cycles of irregular length. Implantable progestin, such as levonorgestrel in silastic capsules (Norplant), has gained in popularity, and monthly injectable preparations, such as Cycloprovera and HPR 102, are useful for certain patients.

DOSAGE OF PROGESTOGEN AND ESTROGEN

Historically, the dose of a progestogen that produced secretory changes in the estrogen-primed endometrium was the one used in the first clinical contraceptive studies of that agent. Because of irregular bleeding associated with use of the progestogens, estrogens were added but in amounts that alone could almost completely block fertility through ovulation inhibition.

Dosage reduction was approached empirically. Once a satisfactory ratio of progestogen/estrogen doses was established, the components were reduced in the same proportion until an unacceptable incidence of pregnancy or irregular bleeding ensued. At that point, the estrogen dose was increased until the product became clinically acceptable, that is, in terms of contraceptive efficacy and regulation of bleeding.

Because competition for lower dose progestogen products was keen, at times shortcuts were taken in selecting these lower doses. In some instances, progestogen doses were decreased to correspond to those of other products simply on the basis of broad assumptions about their relative potency. Estrogen doses, however, rather than undergoing proportional decreases, were retained at sufficiently high levels to ensure contraceptive efficacy of the product. It later became apparent that these levels of estrogen were unnecessarily high when it was learned that the progestogen dose alone provided a fully contraceptive effect. Currently, the dose of estrogen is primarily determined by the amount required to prevent breakthrough bleeding.

In establishing a minimal effective dose, it must be appreciated that doses are not tailored to the individual. Significant interindividual variation exists because of differences in body weight, absorption, tissue receptor levels, metabolism, diet, and, in some cases, drug interactions with either the estrogen or progestogen component. Because the dose must be the minimal effective dose for the entire population (ED100), the dose for some individuals, or at least for some cycles in some individuals, may be in great excess of that required for contraception. The reason for this is that the coefficients of variation of pharmacological parameters are in the range of 3055%. Unexpectedly, the variability from one cycle to the next within individuals is only slightly less than the interindividual variability. Despite this variability, significant differences between population groups do exist,1, 2 and some tailoring of the dose on this level may be appropriate.

PROGESTOGENS

The synthetic progestogens are metabolized and inactivated much more slowly than the natural progesterone. Primarily two types of compounds have been prepared: compounds with or without the angular methyl group in the C-19 position of the steroid molecule (the 17-acetoxy compounds, such as medroxyprogesterone acetate, and the 19-nor compounds, such as norethindrone). Today only the 19-nor compounds are formulated in oral contraceptives (OCs). The 17-acetoxy compounds are used only in injectable contraceptives.

The 19-nor steroids may be considered as modifications or analogs of norethindrone. Levonorgestrel, which is now prepared in the active L form rather than in the D,L form prepared previously, has only one additional methyl group. Instead of a methyl group extending from C-13, it has an ethyl group (see elsewhere in The Global Library of Women's Medicine). Gestodene differs from levonorgestrel in that it has an additional double bond at C-15. Desogestrel differs from levonorgestrel in that it has a methylene group at C-11; it also lacks the 3-ketone group, but that is added rapidly in vivo to form 3-ketodesogestrel. Norgestimate differs from levonorgestrel in that it has an oxime at the C-3 carbon in place of the ketone; it also has an acetate group at C-17 but that group is cleaved rapidly in vivo (the deacetyl norgestimate has a half-life of 1617 h compared to a half-life of about 1 h for norgestimate).

The pharmacokinetics of the 19-nor progestogens are relatively similar. Their unique properties are more related to the specificity of binding to receptors for progesterone, androgens, and glucocorticoids. The parameters are given in Table 1 for the progestogens.3, 4, 5, 6, 7, 8

Table 1. Pharmacokinetic properties for progestogens used in oral contraceptive

Progestogen + EE dose

Dose (mg)

Cmax (ng/ml)

EL (h)

Tmax (h)

AUC (ng/h/ml)

MCR (L/d)

LNG + 30 μg EE4

0.125

 

2.6

 

13

 

1.2

 

 

 

158

 

NET + 120 μg EE5

1

 

16

 

8

 

1.2

 

84

 

531

 

MPA6

trace

--

5.5

2

 

626

NGM + 35 μg EE7

0.18

 

3.6

 

16

 

1.5

 

 

 

 

 

DGL + 30 μg EE8

0.15

 

3.7

 

24

 

1.5

 

36

 

100

 

GDN + 30 μg EE9

0.075

 

3.6

 

12

 

1.7

 

35

 

51

 


LNG, levonorgestrel; NET, norethindrone acetate; MPA, medroxyprogesterone acetate; NGM, norgestimate (deacetyl form); DGL, desogestrel (3-ketodesogestrel form); GDN, gestodene; Cmax, maximum concentration in plasma after a single dose of steroid; T½ EL, half-life of elimination; Tmax= the time at which maximum concentration occurred in plasma after a single dose of the steroid; AUC = area under the concentration with time curve; MCR, metabolic clearance rate.
Units were adjusted from the original publications for uniformity in the table. MCR was calculated in some instances from the original dose and AUC data for a body weight of 58 kg and body surface of 1.66 m2.

The progestogens have variable first pass metabolism by the liver, depending on the binding affinity of the steroid to sex hormone-binding globulin (SHBG). Because of weak SHBG binding, norgestimate has the greatest first pass metabolism among the common progestogens. Although ethinyl estradiol has demonstrable enterohepatic circulation, there is no evidence for recirculation of the progestogens.9


Biological properties of progestogens

Blood levels of the progestogens are determined by the rate of both absorption and metabolism, as well as by binding to plasma proteins. SHBG (testosterone–estradiol binding globulin) differs in its ability to bind several progestogens. Gestodene has a binding affinity for SHBG that is close to that of testosterone. Levonorgestrel has about half the affinity of testosterone; 3-ketodesogestrel and norethindrone have 18% and 10%, respectively. The degree of binding affects the clearance (elimination) of the drug and the concentration of free steroid in blood.

The concentration of free or unbound drug is that which is available to diffuse from the bloodstream into the target organs to bind to receptors in the hypothalamus, pituitary, mammary gland, and uterus. The activity of the drugs may be measured as the affinity for the progesterone receptor. It may also be measured as the ability to suppress luteinizing hormone (LH) secretion or to stimulate endometrial development in rabbits. Levonorgestrel has about a threefold greater affinity than progesterone for the progesterone receptor, and norgestimate has an affinity similar to that of progesterone. Estimation of equivalent doses, however, must take into account not only the activity of the drug at the target cell, but also the absorption of the drug and the residence time of the drug in the body. Since levonorgestrel and norgestimate have similar absorption characteristics and half-lives, the biological activities are proportional to their affinities for the progesterone receptor. The doses of the two progestogens for the LH and endometrial responses correspond to their binding affinities. However, natural progesterone is poorly absorbed and has a much shorter half-life, approximately 20 min, and, therefore, the dose for an equivalent effect is much higher.

Suppression of gonadotropin secretion appears to be primarily limited to the preovulatory surge at midcycle. Tonic levels of gonadotropins at other times during the menstrual cycle have not been found to be greatly suppressed. In a study of the 'minipill' (300 μg of norethindrone daily), Landgren and associates10 found that 40% of subjects had relatively normal patterns of gonadotropins and steroids during the menstrual cycles. Even in cycles in which ovulation did not occur based on serum progesterone levels, estradiol levels often increased to normal preovulatory concentrations but LH concentrations often had bizarre patterns; the synchrony of LH with follicular development appeared to be lost. Since the minipill gave good contraceptive efficacy, it is obvious that mechanisms other than inhibition of ovulation were involved.

As is frequently the case with synthetic steroids, the specificity of the progestogens is less than that of the natural hormones. Levonorgestrel, a derivative of testosterone, has relatively high affinity for the androgen receptor, and stimulates an increase in the rat ventral prostate weight. Relative to dihydrotestosterone, the affinities of progesterone, norgestimate, 3-ketodesogestrel, gestodene, and levonorgestrel for the androgen receptor are 0.003, 0.005, 0.118, 0.154, and 0.220, respectively.11 Undesirable side-effects associated with the androgenic activity of the progestogens include acne and seborrhea, weight gain, blood pressure increase, and an increase in the low-density lipoprotein (LDL) to high-density lipoprotein (HDL) ratio. The 17-acetoxy progestogens, medroxyprogesterone acetate, chlormadinone acetate, and megestrol acetate, also have androgenic activity in high doses,12 but have little androgenic activity in doses used for contraception, as assessed by the degree to which they are capable of suppressing SHBG levels.13 Obviously, the newer progestogens have an advantage with respect to these side-effects. Based on animal studies12 and in human subjects receiving high doses for cancer therapy,14 the 17-acetoxy progestogens also appear to have significant glucocorticoid activity. In addition, the 19-nor steroids evidently are capable of exerting some estrogenic activity since they have been shown to stimulate proliferation of mammary tumor cells only in cells that contain the estrogen receptor.15 The 17-acetoxy progestogens do not have this property.

 

Carcinogenic potential of progestogens

The potential of progestogens in stimulation of breast cancer cell division may be explained by the estrogenic activity of the 19-nor compounds as described above. 17-Acetoxy progestogens such as MPA, however, do not exert demonstrable stimulation of mammary tumor cell growth in vitro,16 and may act to inhibit growth through both progesterone and glucocorticoid receptors. Effects of the combined OCs are considered below.

 

Cardiovascular effects

The primary concern of progestogens is their effects on circulating lipoproteins. In general, the progestogens increase LDL and decrease HDL levels in blood. Increases in the LDL/HDL ratio are associated with atherosclerotic plaque formation and subsequent deleterious circulatory deficits. This effect of progestogens exacerbates the more life-threatening thromboembolic effect of estrogens. The effect is dose-dependent and appears related to the androgenicity of the progestogen, but even medroxyprogesterone acetate induces moderate changes in the LDL/HDL ratio. Nevertheless, with the newer low-dose formulations, the effect of the progestogens is probably not significant, even with a relatively androgenic progestogen such as levonorgestrel.

 

Carbohydrate effects

Although the side-effects of progestogens alone have not been studied as extensively as the combined progestogen and estrogen preparations, it is clear that the progestogens are responsible for the effects on carbohydrate metabolism. Glucose tolerance is impaired by depomedroxyprogesterone acetate administered without estrogen, but the effect is less than that of levonorgestrel combined with ethinyl estradiol (EE).17

 

ESTROGENS

Ethinyl estradiol (EE) was the first synthesized orally active steroidal estrogen. It has been used clinically for almost 40 years. In 1943, it was shown that 0.05 mg EE daily would control the symptoms of dysmenorrhea and presumably ovulation. This concept of ovulation inhibition remained dormant until the synthesis of the 19-nor progestogens, norethindrone and norethynodrel. Both of these compounds were made from estrone and contained mestranol as a not-easily removed contaminant. As a result, mestranol was used as the estrogen in the original norethindrone and norethynodrel combinations. Improvements in the chemical process facilitated development of estrogen-free norethindrone, and it became possible to use EE in place of mestranol. In vivo, the methyl group is removed from mestranol, and the biologically active compound that binds to the estrogen receptor is EE.

Although the variability among individuals is great, it has been established that the effective conversion of mestranol to EE in the human is approximately 70%.18 Therefore, a dose of 50 μg of mestranol is equivalent to 35 μg of EE. Moxestrol (11β-hydroxy-17-ethinyl-estradiol) is 10-times more potent than EE but has not been used in OC preparations. However, it may have an advantage over EE in that it is not metabolized to catechols or epoxides, and has less potential for inducing neoplastic transformation of its target cells.18

EE has an enterohepatic circulation. That is, it is excreted as conjugates of glucuronic and sulfuric acid in the bile. The conjugates are hydrolyzed to a significant extent by microorganisms in the digestive tract, and the EE is then reabsorbed. This increases the mean residence time of EE in the body. Although it is possible that a change in the flora of the gut may increase or decrease the hydrolysis of the conjugates and thereby the reabsorption of EE, in general, alteration of the flora of the gut by antibiotics does not affect the availability of EE.

EE has an initial (distribution) half-life (T1/2) of 0.52.4 h and an elimination T1/2 of 13.127 h. The time for maximal serum concentrations (Tmax) is 12 h.18

ESTROGEN-PROGESTOGEN COMBINED PREPARATIONS

Dose requirements and effective ratios of estrogen to progestogen

Estrogens and progestogens are antagonistic at the level of the uterus and vagina. In the breast, however, progesterone does not oppose the actions of estrogens, but directs the proliferative effects of prolactin and growth hormone toward the development of the lobulo-alveolar system. Except for androgenic and estrogenic activities of the synthetic compounds, the progestogens presumably act similarly. At the hypothalamic–pituitary level, there also appear to be more complementary actions of estrogens and progestogens.

To demonstrate the interaction of progestogen and estrogen at the hypothalamic–pituitary level, norethindrone, 0.4 mg, was combined with EE, 0.035 mg. At these respective doses, norethindrone fully suppressed fertility but inhibited ovulation about 70% of the time and EE inhibited progestogen-induced bleeding irregularity but was not fully effective alone in blocking ovulation. Together, the two agents provided total antiovulatory activity.19

Thus, there is an overlap of activities that produce a synergistic response with this and other progestogen/estrogen mixtures. It implies that higher doses of either or both components are able to create a degree of hypothalamic–pituitary suppression that exceeds the therapeutic need for ovulation inhibition. It follows that an OC combining a progestogen at its ED100 antifertility dose with an estrogen in doses sufficient to control irregular cyclic bleeding will be as effective as higher dose products. This has proven to be the case.


Biological effects of combined contraceptive steroids

Measurements of LH and follicle-stimulating hormone in serum or urine have revealed that the higher doses of OCs (2 mg norethindrone + 100 μg mestranol) only partially suppress tonic serum follicle-stimulating hormone (FSH) and LH levels, but effectively eliminate the midcycle peak of gonadotropin secretion. Lower doses of norethindrone and mestranol (1 mg + 50 μg, respectively) were also found to suppress the preovulatory peak of gonadotropin secretion, but endogenous estradiol excretion was not completely suppressed; about 20% of the time normal preovulatory increases in estradiol occurred.20 Unfortunately, few such studies have been conducted with the newer, low-dose agents. In one study, Corson21 studied hormonal responses to norgestimate and EE (250 μg and 35 μg, respectively) and levonorgestrel and EE (300 μg and 30 μg, respectively). Serum progesterone levels remained uniformly low as expected, but serum SHBG levels varied with the preparation. Since serum SHBG concentrations have been shown to be suppressed by androgen and increased by estrogens, the level of serum SHBG provides a good bioassay for the overall effect of the combined steroids. The preparation with levonorgestrel and EE increased SHBG levels by about 25%. Administration of norgestimate and EE resulted in an approximately 100% increase in SHBG, at least in part because of the lower androgenicity of norgestimate. Nevertheless, the contribution of serum estradiol to the levels of SHBG and other estrogenic responses remain to be determined.

Monthly injectable preparations of Cycloprovera and estradiol cypionate (25 mg and 5 mg, respectively) or HPR 102 (50 mg norethindrone enanthate and 5 mg estradiol valerate) provide suppression of ovulation for approximately 60 days and suppression of follicular development for a minimum of 30 days. Estradiol is initially elevated in the range of 12 nmol/L as a result of hydrolysis of the cypionate or valerate esters. As the drugs are cleared from the body, there is an increase, after 3040 days, in endogenous estradiol to 12 nmol/L, which lasts for at least 10 days before a menstrual period occurs.

A newer mode of administration has been popularized for reducing menstrual periods to four times per year. Cycles are extended by Seasonale, composed of levonorgestrel and ethinyl estradiol (Barr Laboratories, Inc.) and generic forms by eliminating two drug-free intervals from the usual monthly regimen.22 The discontinuation rate in a phase III clinical trial was 42%, with 9.5% discontinuing because of adverse events. Bleeding or spotting averaged 8 days for the first cycle and 4 days in the eighth cycle.


Cardiovascular effects

An association between and an increase in thromboembolic disease and the use of combined OCs has been demonstrated by several older epidemiologic studies. The evidence suggests that the incidence of venous thromboembolism correlates with the EE dosage and the arterial complications with the type and dose of progestogen. Estrogen has a dose-dependent suppressive effect on natural inhibitors of clotting factors, including antithrombin III and proteins C and S. The progestogens affect primarily the lipoproteins, increasing LDL and decreasing HDL. However, the low-dose preparations with less androgenic progestogens have little or no significant effect on serum lipoprotein levels since the estrogenic activity, which decreases the LDL/HDL ratio of the preparations, are dominant. The effect of the progestogens on carbohydrate metabolism also is a risk factor for vascular disease because impaired glucose tolerance raises serum insulin levels and results in the associated angiopathy. However, the newer epidemiologic studies now indicate that the incidence of cardiovascular disease with the low dose current OCs has been substantially reduced. In part, this is due to better drugs and, in part, to a better selection of patients for steroidal contraception. A metaanalysis of newer combination pills has concluded that there is no significant association of OCs with cardiovascular disease regardless of duration of use.23


Cancer

There appears to be little or no increased risk of breast cancer associated with OC use. The relative risk in the WHO studies was 1.2 in developing countries and no increased risk in developed countries.24 One concern, however, is that women who used OCs for 512 years had a relatively greater risk (1.42–1.46) of developing breast cancer before the age of 45.25, 26 Another concern is whether OC use in women with a history of benign breast disease promotes development of malignant disease in premenopausal women.

Cervical cancer is difficult to attribute to the use of OCs since the effects of confounding factors are difficult or impossible to eliminate. Nevertheless, the WHO report concluded that the use of OCs for more than 5 years was associated with a relative risk of cervical cancer from 1.3 to 1.8.

Endometrial cancer risk is lower in women who have used oral contraception. Relative risks as low as 0.5 have been reported.

Ovarian cancer is reduced by approximately 50% in women who used oral contraceptives but those with BRCA1/2 mutations have no advantage.27


Drug interactions

Specific pharmacologic interactions between OCs and other drugs have been shown to increase or decrease the effectiveness of the contraceptive agents.28 In addition, there are a number of medications that are affected by the ingestion of OCs. Drugs may reduce the efficacy of OCs by induction of liver enzymes that metabolize steroid hormones. They may also compete for SHBG binding or elevate the concentration of serum SHBG. Other drugs may interfere with intestinal absorption or the enterohepatic circulation.

Factors known to induce liver enzymes of steroid metabolism are the antibiotics: rifampin, penicillin, chloramphenicol, cephalosporins, metronidazole, and tetracyclines; the sulfonamides; nitrofurantoin; the anticonvulsants: phenobarbital, primidone, carbamazepine, ethosuximide, and phenytoin; and the antifungal griseofulvin. The list is not comprehensive but is an indication of the number and types of compounds that may affect activity of the OCs. By inducing enzymes that hydroxylate steroids in the liver, these compounds increase the clearance and thereby decrease the average blood levels of the OCs.

Potentially, a change in the intestinal flora, for example by antibiotic medication, may decrease the enterohepatic circulation of EE by decreasing the hydrolysis of glucuronide and sulfate conjugates of the estrogen. This results in less unconjugated EE that can be reabsorbed into the bloodstream; lower serum levels of EE; and a higher proportion of conjugated steroid that is readily excreted.

Plasma levels of EE may be increased by some drugs. Ascorbic acid and paracetamol (acetaminophen) compete with EE for sulfation. With less sulfate conjugates formed, more EE is available for reabsorption from the gut and serum levels are increased.

On the other hand, it should be realized that the drugs that decrease the efficacy of OCs compete for the same enzyme systems in the liver that metabolize the OCs. Such drugs may, therefore, be metabolized more slowly in the presence of OCs, resulting in elevated blood levels of the drugs. The presence of OCs may, therefore, decrease the rate at which the antibiotics, anticonvulsants, and antifungals are metabolized, and may increase their concentrations to toxic levels. Other drugs which themselves are not good inducers of liver enzymes may also be affected similarly; antidepressants, beta blocking agents, some antianxiety agents, and theophylline preparations may be increased in blood by OCs. The opposite effect, that of decreased efficacy, may result from concomitant administration of OCs and other drugs: EE results in decreased efficacy of anticoagulants; the progestogens decrease the effectiveness of insulin and anti-diabetogenic agents.

CONCLUSIONS

The steroidal contraceptives have been greatly improved over time. The 19-nor progestogens have good pharmacokinetic properties. They are efficiently absorbed, have long biological half-lives, and have high affinity for the progesterone receptor. Some compounds had undesirable side-effects, which were primarily related to the androgenic activity they possessed. Such side-effects as weight gain, blood pressure increase, undesirable lipoprotein profiles, and diabetogenic properties have been largely eliminated by exceedingly small modifications in the levonorgestrel molecule and by the use of smaller doses. The problem of determining the minimal effective dose is one that could only be determined with considerable experience in clinical investigations and could not be ascertained from preclinical animal studies. The estrogenic component remains an essential part of the OCs to control menstrual and intermenstrual bleeding. Mestranol has largely been replaced by EE, although EE is the active product of both. The thromboembolic disease associated with OCs is primarily due to the estrogen component. This has been minimized by decreasing the dose of EE but remains the primary concern of OC users. It is clear that steroidal contraceptives decrease the incidence of endometrial and ovarian cancer, and the possible increase in the incidence of breast and cervical cancer is too small to ascertain with certainty.

REFERENCES

1

Goldzieher JW, Dozier TS, de la Pena A. Plasma levels and pharmacokinetics of ethynyl estrogens in various populations. I. Ethynyl estradiol. Contraception 1980;21:1

2

Fotherby K, Akpoviroro J, Abdel-Rahman HA et al. Pharmacokinetics of ethynyloestradiol in women from different populations. Contraception 1981;23:487

3

Kuhnz W. Pharmacokinetics of the contraceptive steroids levonorgestrel and gestodene after single and multiple oral administration to women. Am J Obstet Gynecol 1990;163(suppl):2120

4

Stanczyk FZ. Pharmacology, pharmacokinetics and pharmacodynamics: Pharmacokinetics of progestogens. Int Proc J 1989;1:11

5

Utaaker E, Lundgren S, Kvinnsland S, Aakvaag A. Pharmacokinetics and metabolism of medroxyprogesterone acetate in patients with advanced breast cancer. J Steroid Biochem 1988;31:437

6

McGuire JL, Phillips A, Hahn DW et al. Pharmacologic and pharmacokinetic characteristics of norgestimate and its metabolites. Am J Obstet Gynecol 1990;163:2127

7

Bergink W, Assendorp R, Kloosterboer L et al. Serum pharmacokinetics of orally administered desogestrel and binding of contraceptive progestins to sex hormone-binding globulin. Am J Obstet Gynecol 1990;163:2132

8

Tauber U, Tack JW, Matthes H. Single dose pharmacokinetics of gestodene in women after intravenous and oral administration. Contraception 1989;40: 461

9

Orme ML, Back DJ. Factors affecting the enterohepatic circulation of oral contraceptive steroids. Am J Obstet Gynecol 1990;163:2146

10

Landgren BM, Lager S, Diczfalusy E. Hormonal effects of the 300 &b.mu;g norethistrone (NET) minipill. Contraception 1981;23:269

11

Phillips A. The selectivity of a new progestin. Acta Obstet Gynecol Scand 1991;1(suppl 125)1:45

12

Labrie C, Cusan L, Plante M et al. Analysis of the androgenic activity of synthetic “progestins” currently used for the treatment of prostate cancer. J Steroid Biochem 1987;28:379

13

Pedersen OD, Jensen HK: Long-acting treatment with transcutaneous estradiol and oral medroxyprogesterone acetate. Acta Obstet Gynecol Scand 1992;71:593

14

Blossey HC, Wander HE, Koebberling J, Nagel GA. Pharmacokinetic and pharmacodynamic basis for the treatment of metastatic breast cancer with high-dose medroxyprogesterone acetate. Cancer 1984;54:1208

15

Jordan VC, Jeng MH, Catherino WH, Parker CJ. The estrogenic activity of synthetic progestins used in oral contraceptives. Cancer 1993;71(suppl):1501

16

Sutherland RL, Hall RE, Pang GY et al. Effect of medroxyprogesterone acetate on proliferation and cell cycle kinetics of human mammary carcinoma cells. Cancer Res 1988;48:5084

17

Liew DF, Ng CS, Heng SH, Ratnam SS. A comparative study of the metabolic effects of injectable and oral contraceptives. Contraception 1986;33:385

18

Goldzieher JW. Selected aspects of the pharmacokinetics and metabolism of ethinyl estrogens and their clinical implications. Am J Obstet Gynecol 1990;163:318

19

Rudel HW, Kincl FA. Novel oral contraception combination. U S Patent 3,836,651, September 17, 1974

20

Elstein M, Briston PG, Jenkins M et al. Effects of a low-oestrogen oral contraceptive on urinary excretion of luteinizing hormone and ovarian steroids. Br J Med 1974;1:11

21

Corson SL. Contraceptive efficacy of a monophasic oral contraceptive containing desogestrel. Am J Obstet Gynecol 1993;1687:1017

22

Anderson FD, Gibbons W, Portman D. Long-term safety of an extended-cycle oral contraceptive (Seasonale): a 2-year multicenter open-label extension trial. Am J Obstet Gynecol 2006;195(1):92-6

23

Chasan-Taber L, Stampfer MJ. Epidemiology of oral contraceptives and cardiovascular disease. Ann Intern Med 1998;128(6):467-77

24

WHO Task Force on Oral Contraceptives. A multicentre study of coagulation and haemostatic variables during oral contraception: Variations with four formulations. Br J Obstet Gynaecol 1991;98:1117

25

Thomas DB. Oral contraceptives and breast cancer: Review of the epidemiological literature. Contraception 1991;43: 597

26

Romieu I, Berlin JA, Colditz G. Oral contraceptives and breast cancer. Review and meta-analysis. Cancer 1990;66:2253

27

Modan B, Hartge P, Hirsh-Yechezkel G, Chetrit A, Lubin F, Beller U, et al. Parity, oral contraceptives, and the risk of ovarian cancer among carriers and noncarriers of a BRCA1 or BRCA2 mutation. N Engl J Med 2001;345(4):235-40

28

Angle M, Huff PS, Lea JW. Interactions between oral contraceptives and therapeutic drugs. Outlook 1991;9:1