HIGH-DOSE ESTROGEN THERAPY—THE FIRST FOOD AND DRUG ADMINISTRATION-APPROVED REGIMEN
Although estrogens from ovarian extracts were originally used in the 1920s for their ability to prevent pregnancy,5 the first study of high-dose oral estrogen for emergency postcoital contraception was published in 1968 by Haspels and the International Planned Parenthood.6 Although many of the early studies used diethylstilbestrol (DES), other equivalent estrogen regimens advocated ethinyl estradiol, conjugated estrogens, and estrone (Table 1). When initiated within 72 hours of unprotected intercourse, therapy with DES had a reported failure rate of only 0.0% to 2.4%.7 Failure rates were similarly low with other high-dose estrogen regimens. In a review of four published studies using ethinyl estradiol at daily doses between 2 and 5 mg for 5 days, Fasoli and coworkers11 calculated an aggregate pregnancy rate of 0.6%, and a failure rate of 1.6% was reported with the conjugated estrogens.12
Table 1. Dosing Regimens for Diethylstilbesterol (DES) and Other Types of Estrogen-Only Emergency Contraception
25–50 mg bid × 5 days
2.5 mg bid × 5 days
10 mg bid × 5 days
5 mg bid × 5 days
bid, twice daily.
However, troublesome side effects made this form of emergency contraception particularly unpleasant. In a study of 3016 women treated with DES, nausea was reported in 53% of women and vomiting in 21%.7 Indeed, nausea was so common with this regimen that many physicians routinely prescribed an antiemetic to be taken with the first DES dose. Other reported side effects included the occurrence of menorrhagia with the next menses in 20%, mastalgia in 10%, and occasional headaches and dizziness.7 Compliance with this regimen was often difficult, mainly because of the significant side effects and the 5 required days of treatment.
The use of DES for postcoital contraception was approved by the FDA in 1975. However, as physician and patient awareness increased regarding the relation between fetal DES exposure and vaginal adenocarcinoma and genital malformation or dysfunction in both female and male offspring, this regimen fell into disfavor. The FDA has since discontinued its approval of DES for postcoital contraception.13
PREVEN AND THE YUZPE METHOD: COMBINED ESTROGEN AND PROGESTIN CONTRACEPTIVE PILLS
After the loss of FDA approval for the postcoital contraception indication for DES, there was no officially sanctioned emergency contraception in the United States until Preven (Gynétics, Sommerville, NJ) was approved in 1998. In the intervening years, the off-label use of oral contraceptive pills, known as the Yuzpe method, was popularized. This regimen consists of two oral contraception pills, each containing ethinyl estradiol 50 μg and dl-norgestrel 0.5 mg, administered within 72 hours of unprotected intercourse with a second dose repeated 12 hours later. This provides a total dosage of 200 μg ethinyl estradiol and 2 mg dl-norgestrel. To prescribe the Yuzpe regimen, physicians and pharmacists had to break apart an Ovral monthly packet and repackage four tablets. A prescription for Preven provides equivalent medication (four tablets) with patient instructions and a urine pregnancy test in an individually packaged kit. The only difference between a Preven tablet and Ovral is that the dl-norgestrel 0.5 mg has been replaced by levonorgestrel 0.25 mg. Because the levonorgestrel enantiomer is the bioactive component of dl-norgestrel, the two products are pharmacodynamically equivalent. Therefore, results of research with the Yuzpe regimen are applicable to current use of Preven.
After preliminary studies in 197414 and 1977,15 Yuzpe published his seminal multicenter Canadian report advocating postcoital use of combined ethinyl estradiol/dl-norgestrel oral contraceptive pills in 1982.16 Numerous confirmatory studies followed, and this regimen rapidly replaced DES and other high-dose estrogens as the most commonly prescribed form of emergency contraceptive. 17,18,19 Compared to high-dose estrogen therapy, the Yuzpe method offered a 125-fold reduction in ethinyl estradiol content and a decrease in the overall length of drug therapy.
A particular appeal of the Yuzpe method was that it used readily available agents that enjoyed high physician and public confidence. Although Ovral was probably the most frequently prescribed medication for emergency contraception, other less-studied alternative regimens available in the United States included the following20: (1) four oral contraceptive tablets, each containing ethinyl estradiol 30 μg and levonorgestrel 0.125 mg repeated 12 hours later (possible medications: the yellow tablets [third week] of either Tri-Levlen or Triphasil) or (2) four oral contraceptive tablets each containing 30 μg of ethinyl estradiol and 0.15 mg of levonorgestrel (or dl-norgestrel 0.3 mg), repeated 12 hours later (possible medications: Nordette, Levlen, or Lo/Ovral).
In a pooled analysis of 24 studies using the Yuzpe method in 9588 women, the calculated average failure rate was 1.84% (range, 0.0% to 7.44%).10 The true effectiveness of this method in preventing pregnancy is difficult to determine and can only be estimated on the basis of predicted pregnancy rates. The risk of pregnancy with one act of intercourse ranges from 0.0% to 26% depending on the day of intercourse relative to ovulation. The risk is highest during the 3 days before ovulation and on the day of ovulation itself. During this window, estimated conception rates range from 15% to 26%.21,22 Reviewing studies based on the use of postcoital oral contraceptives, Trussell and Stewart23 calculated that the Yuzpe method could reduce the risk of pregnancy by 75% and would have a failure rate of approximately 25%. Garcia and associates24 hypothesized that failures of emergency contraception occurred because: (1) the woman had multiple sexual exposures during the month, (2) a pregnancy was already established at the time of administration, (3) inadequate medication was prescribed, or (4) the woman did not take the pills as prescribed or discontinued them because of nausea and vomiting.
There is some controversy among authors as to the importance of the timing of initiation of emergency oral contraceptive therapy. Most authors advocate starting the first dose of the Yuzpe method within 72 hours of unprotected intercourse. Beginning therapy as soon as possible, even within that window, may also affect pregnancy rates. Kane and Sparrow25 found that women who started treatment within 12 to 24 hours of unprotected intercourse had improved success rates when compared to women who delayed therapy until 48 to 72 hours. Improved success rates with early administration were also seen in the 1993 study by Ho and Kwan26 comparing levonorgestrel with the Yuzpe method as well as in the World Health Organization's trial with the same medications published in 1998.27
Conversely, in a review of nine studies that included documentation of time from unprotected intercourse to initiation of therapy, Trussell and colleagues20 found that delaying the onset of treatment to day 3 did not affect outcome, including a delay between 72 and 120 hours in some women. As a result, they and other authors recommended that emergency oral contraception therapy be considered even beyond day 3.20,28 However, the American College of Obstetrics and Gynecology in their publication, Practice Patterns: Emergency Oral Contraception 1996,29 thought that there were insufficient data to evaluate the effectiveness of initiating emergency contraception after 72 hours.
Nausea and vomiting, although less bothersome than with high-dose estrogen therapy, are still common side effects with this regimen. Nausea, which may occur after either dose and last for as long as 2 days, has been reported to occur in as many as 30% to 60% of women with emesis in up to 22%.16,30,31 Taking the medication with food or meals often reduces the nausea. Because antiemetics do not appear to be as effective if they are taken after the onset of symptoms,19 women are often encouraged to take them 1 hour before each dose.32,33 Some physicians and clinics routinely prescribe antiemetics, thus eliminating the patient's need to recontact her provider.29
Occasionally, patients are concerned that they may have vomited their contraceptive pills if they have an episode of emesis shortly after taking the medication. There are no published data associating early onset of emesis with an increased risk of failure; however, none of the studies were designed to evaluate this effect specifically. One may postulate that if the underlying cause of the nausea is from an estrogen-mediated effect on the central nervous system, sufficient absorption of the estrogen would have occurred by the time the patient began to experience these symptoms.29 Some authors have encouraged patients to take an extra set of oral contraceptive pills should emesis occur within 2 to 3 hours of ingesting a dose. However, taking additional medication does not increase the efficacy of the therapy and may only worsen the centrally mediated nausea. Other less-common reported side effects include abdominal pain (2.1% of women treated), dizziness (2.1%), uterine cramping (1.4%), and, rarely, breast tenderness (1% to 4%), headache, and fluid retention.16
The contraceptive effect of postcoital birth control pills is thought to be mediated through a number of different mechanisms, including the following: (1) changes in cervical mucus that reduce sperm penetration; (2) suppression of pituitary hormone secretion, which inhibits or delays ovulation; (3) disruption of luteal function by direct action on the corpus luteum; (4) alteration of tubal gamete transport; (5) suppression of endometrial estrogen and progesterone receptors when administered within 5 days of the luteinizing hormone (LH) surge; and (6) modification of endometrial development, which interferes with implantation of a potentially fertilized ovum.34,35,36,37,38,39,40,41 Endometrial biopsies performed after postcoital oral contraception show a marked asynchrony in the maturation of the endometrial glands and stroma. This disordered, out-ofphase endometrium interferes with implantation.14 The particular mechanism depends on the precise point within a menstrual cycle in which postcoital hormonal therapy is administered. If given before ovulation, follicular development and ovulation are altered; with postovulatory administration, the effect on the endometrium predominates.
Administration of postcoital oral contraceptives may affect the onset of the woman's next menstrual period.19,40,41,42 Timing of the postmedication withdrawal bleed depends on the exact phase of the woman's cycle during which postcoital contraceptives are administered. If she takes the pills in the periovulatory period, the next menses usually begins a few days earlier than expected. If exposure occurs in the follicular phase, the cycle is aborted and the next menses may begin as early as day 21. Luteal phase administration may delay the onset of the menstrual bleed by a couple of days. The duration of the menstrual flow is normal in more than 90% of women.30 The effect on menstrual cycle appears to be independent of the woman's history of regular or irregular cycles.15,16 Because 98% of women menstruate by 21 days after treatment, it is important to advise the woman that she may be pregnant if her next menstrual period does not begin within that period.15,16,25,33 She should be encouraged to obtain a pregnancy test and seek medical evaluation and counseling.
No published studies have reported evidenced-based criteria that would contraindicate the use of combination hormonal emergency contraception.29 Although some studies have excluded women with absolute contraindications to oral contraceptives, there have been no reports of major cardiovascular or neurologic adverse events in any women. A history of thromboembolism is considered only a relative contraindication to hormonal postcoital contraception. A review of the records of 73,302 women who received emergency contraception in the United Kingdom between 1989 and 1996 identified 19 women in whom venous thromboembolism developed; however, none of them had been treated within 45 days of the diagnosis.43 Although the daily dose of steroid hormones in emergency contraception is greater than that in normal oral contraceptive use, the duration is very short. It is thought that the risk of oral contraceptive hormones is based on long-term exposure and is not likely to pertain to emergency contraception.19,25,44
PLAN B AND PROGESTIN-ONLY CONTRACEPTIVE PILLS
In 1999 the FDA approved the first progestin-only emergency contraception pill, Plan B, manufactured by Women's Capital Corporation (Bellevue, WA). A prescription for Plan B provides two pills, each containing levonorgestrel 0.75 mg, with instructions to take one pill as soon as possible within 72 hours after unprotected coitus and the second pill 12 hours later.
Although this alternative emergency contraception regimen had been known to physicians in the United States, before the approval of Plan B, it was little used because of the lack of a convenient progestin-only pill. Two different regimens had been studied. The most popular one required administration of 0.75 mg levonorgestrel beginning no later than 8 hours after intercourse with a repeat dose 24 hours later. If another episode of unprotected intercourse occurred on a subsequent day, it was recommended that the two levonorgestrel doses be repeated until seven doses had been administered.45,46 Some authors suggested that the initial levonorgestrel dose may be started as late as 48 hours after unprotected intercourse.45 An alternate, less-studied regimen consisted of a single dose of 0.6 mg levonorgestrel administered within 12 hours of intercourse.47 Before 1999 levonorgestrel-only tablets were not available in the United States. Instead, an equivalent dosage of the minipill, Ovrette, was prescribed. One Ovrette tablet contains 0.075 mg norgestrel, which is equivalent to 0.0375 mg levonorgestrel. Using the available medication, progestin-only emergency contraception in the United States required two doses, each consisting of 20 tablets of Ovrette for the first regimen (total of 40 pills) or 16 tablets of Ovrette for the second regimen (total of 32 pills). Development and approval of Plan B has dramatically changed our options for a progestin-only emergency contraception.
FDA approval was based on a large study, published in 1998 by the World Health Organization Task Force on Postovulatory Methods of Fertility Regulation, which showed the levonorgestrel regimen to be more effective and better tolerated than the Yuzpe method.48 In 1955, evaluable women were randomly assigned to the two different regimens; the pregnancy rate in patients who received levonorgestrel was 1.1% versus 3.2% in the women treated with the Yuzpe method. The crude relative risk of pregnancy for levonorgestrel compared with the Yuzpe regimen was 0.36 (95% confidence interval [CI], 0.18, 0.70) and the proportion of pregnancies prevented compared to the expected number without treatment was 85% (95% CI, 74, 93%) with levonorgestrel and 57% (95% CI, 39, 71%) with the Yuzpe method. Nausea (23.1% versus 50.5%) and vomiting (5.6% versus 18.8%) were significantly less frequent with the levonorgestrel regimen than with the Yuzpe regimen (p < .01), and the efficacy of both treatments declined with increasing time since unprotected coitus (p = .01).
The efficacy of the Yuzpe method in the 1998 WHO study was not as good as that reported in previous studies. In their comparison of levonorgestrel and the Yuzpe method published in 1993, Ho and Kwan46 reported a corrected failure rate of 2.6% with progestin-only therapy and 2.4% with the combined oral contraceptives, and as cited above, a pooled analysis of 24 studies using the Yuzpe method in 9588 women calculated the average failure rate was 1.84% (range, 0.0% to 7.44%).10
Because no estrogen is administered in the progestin-only regimen for emergency contraception, gastrointestinal side effects are significantly less common. Similar to the WHO study, which reported reduced complaints of nausea and vomiting in the levonorgestrel-treated women,48 He and colleagues49 also found a low incidence of these symptoms in a study of 361 women treated with levonorgestrel for postcoital contraception. Nausea was reported by 7% of women and only 1% reported emesis. The same study also cited a 1% to 4% incidence of dizziness, abdominal pain, breast tenderness, sleepiness, vaginal discharge changes, and fatigue in women who received levonorgestrel. Irregular bleeding may occur, however, especially if additional doses are prescribed for subsequent unprotected coitus. The onset of the subsequent menstrual flow may be affected as well. In a WHO study published in 1987, 12.5% of women treated had a cycle length of fewer than 20 days or more than 35 days.45
The mechanisms of action of progestin emergency contraception are probably analogous to those of the progestin-only minipill: the endometrium becomes disordered and unsuitable for implantation, cervical mucus is thickened and hostile to sperm penetration, and the LH surge may be inhibited or delayed, albeit inconsistently, and the length of the luteal phase may be significantly reduced.50
Antiprogestins, specifically mifepristone, have also been investigated for their emergency contraceptive potential. Mifepristone 200 mg is currently available in the United States as Mifeprex and has been approved by the FDA in combination with misoprostol for first-trimester pregnancy termination. Initial results of studies comparing mifepristone to other regimens suggest that the antiprogestin may be even more effective and have fewer side effects than combined oral contraceptive pills. In 1991, Webb51 randomly assigned 215 women requesting emergency contraception to receive either 600 mg oral mifepristone, two doses of 600 mg of danazol, or two doses of ethinyl estradiol 100 μg/levonorgestrel 500 mg. Failure rates for all three methods were similar, ranging from 1.85% to 3.7%; however, significant differences were reported in the frequency of nausea and vomiting. Seventy-four percent of women receiving the oral contraceptive pills reported nausea. In comparison, only 36.4% of the women receiving mifepristone and 31.6% of the women treated with danazol experienced nausea. In 1992, Webb and colleagues52 published a second study using the same three regimens with 616 women and showed postcoital mifepristone to be significantly better at preventing pregnancy than either danazol or combined oral contraceptive pills, with pregnancy rates of 0%, 4.66%, and 2.62%, respectively. In a study published the same month, Glasier and associates31 reported another 0% pregnancy rate for 402 women treated with 600 mg mifepristone compared with a 1.01% pregnancy rate for women who received two doses of ethinyl estradiol 100 μg/norgestrel 1 mg. Moreover, women in Glasier's study who received mifepristone experienced significantly less nausea and vomiting than did women treated with oral contraceptive pills. More recently in 1999, WHO53 sponsored a study comparing three different single doses of mifepristone as emergency contraception given within 120 hours of unprotected coitus: 559 women received 600 mg, the standard dose used for first-trimester abortion; 560 women were administered 50 mg; and 565 women were treated with a single 10-mg dose. The pregnancy rates did not differ among the three groups: 1.3%, 1.1%, and 1.2%, respectively. There were no effects of treatment delay on pregnancy rates, and no major side effects were reported.
The onset of the next menstrual period was delayed in approximately 40% of women treated with mifepristone in Glasier's study.31 In the WHO study,53 women also experienced a delay of more than 7 days in the onset of their next menstrual period, with the larger dose (600 mg) being associated with an increased proportion of women reporting a delay (36%) compared to only 18% in women who received the 10-mg dose. This may be disconcerting to some women as they wait for the clinical outcome of their emergency contraception; however, it is easy to provide reassurance with today's sensitive pregnancy tests. The explanation for this delay is found in the extensive literature exploring the effect of mifepristone administration during the different phases of the menstrual cycle.54 Mifepristone exposure during the late follicular phase suspends follicular development, produces a fall in estradiol levels, and prevents the gonadotropin surge. Once antiprogestin levels subside, follicular growth resumes and the cycle continues, albeit longer.55 Midluteal administration of an antiprogestin inhibits LH secretion; ovarian hormonal production falls, and blockage of progesterone receptors within the endometrium produces a menstrual flow. Early luteal mifepristone produces an endometrium that is disordered, immature, and not supportive of implantation.56
The use of danazol as a hormonal emergency contraceptive is limited to historical interest. In the 1980s, when alternative regimens were being investigated, several authors examined the postcoital use of danazol, the isoxazole derivative of 17α-ethinyl-testosterone. Several different regimens have undergone limited study: two sequential oral doses of danazol 400 or 600 mg 12 hours apart, and three sequential oral doses of danazol 400 mg 12 hours apart.30,52,57 In 1983, Rowlands and associates30 treated 101 women with either postcoital danazol or combination oral contraceptive pills. They reported that the women who received danazol were six times less likely than the other women to experience nausea, and none were vomiting. In another study, Webb and coworkers52 randomly assigned 616 women to receive either combination oral contraceptive pills, mifepristone, or danazol. Seventy percent of the women who received the contraceptive pills reported nausea compared with only 30% of the women treated with danazol. Although most authors' failure rates with danazol therapy are comparable to those of the oral contraceptive regimen, Webb's findings suggest that danazol treatment is somewhat less effective. In his study, the pregnancy rate for women receiving contraceptive pills was 2.62% versus 4.66% for women treated with danazol.
The mechanisms by which postcoital danazol prevents pregnancy are unclear. Analysis of pituitary and ovarian hormone levels after danazol therapy showed no consistent changes.58 Instead, the primary site of action is thought to be the endometrium, where alterations in metabolism and steroid receptor concentrations are seen.37 Danazol is contraindicated in women with porphyria or impaired liver, kidney, or heart function. Although the risk of androgenic effects on the fetus is extremely small given the timing of exposure, danazol should not be used in women who would not have an abortion should emergency contraception fail.
The IUD is one of the least-prescribed methods of emergency postcoital contraception and, simultaneously, is one of the most effective ways of preventing an unplanned pregnancy. In 1976, Lippes and colleagues59 published the first description of a copper-containing IUD for emergency contraception. In a review of nine studies using the copper IUD, Fasoli and associates60 reported a pregnancy rate of 0.1%; only one of the 879 women who had received a postcoital IUD conceived. The copper IUD, which can be inserted up to 5 to 7 days after ovulation, alters the endometrium by causing a sterile inflammatory reaction. These changes interfere with sperm transport and produce a decidua that is unfavorable for implantation. The same concerns that limit use of the IUD in women as a primary form of contraception apply to its emergency use. Because of the risk of exacerbating a sexually transmitted infection, women who are just beginning a new relationship, who have been raped, or who have multiple partners are not candidates for postcoital IUD insertion.61 Another important limitation of IUD use is cost; it is clearly the most expensive form of emergency contraception. One justification for its cost, however, is that once placed, the IUD continues to provide long-term contraception for up to 10 years.
“AMERICA'S BEST-KEPT CONTRACEPTIVE SECRET” AND THE OVER-THE-COUNTER DEBATE
Despite the recent FDA approval of Preven and Plan B, 3 decades after Haspels' first publication6 described the use of high-dose estrogen as a “Morning-After Pill,” emergency postcoital contraception remains relatively unknown in the United States among both physicians and sexually active women. In 1995, a sampling of 550 undergraduate and graduate students at Princeton University showed that although most respondents were aware of the existence of postcoital contraceptive methods, 52% confused the combined birth control pill regimen with the use of RU 486 (mifepristone) as an early pregnancy abortifacient.62 To make matters worse, even when women are familiar with emergency contraception, they may not know how to obtain it in a timely fashion.63 A 1994 survey of physicians showed that only 10% of American clinicians informed their patients about the availability of emergency contraception.3 Some of this reluctance may arise because physicians themselves are unsure of the specifics of the different regimens.63
From a practical point of view, initiation of therapy within this 72-hour window is often problematic. If unprotected intercourse occurs at the beginning of a weekend, holiday, or vacation, contacting a physician may be difficult. Even during the week, obtaining a timely appointment may not be easy. At its January 1996 meeting, the Association of Reproductive Health Professionals proposed that an office visit not be considered necessary if the patient is known to the physician or the clinic, a good history is obtained on the phone, and the healthcare provider adequately counsels her before prescribing therapy. The association also suggested that, if indicated, a blood pressure check could be performed at the pharmacy when the patient receives her prescription.10 Others have advocated providing a patient with a “prophylactic” prescription at her annual examination, so that if the need for emergency contraception occurs, the patient can readily obtain therapy in a more timely fashion.
Accessibility would be even further facilitated were the United States to adopt an over-the-counter distribution of hormonal postcoital contraception. Levonorgestrel has been available for this indication in France without a prescription since 1999; in 2000, Norway approved emergency contraceptive pills for over-the-counter purchase, and pills may soon be similarly available within the United Kingdom. A petition to the FDA submitted in February 2001 to consider allowing over-the-counter purchase of postcoital contraception is currently under debate within the American College of Obstetrics and Gynecology and other national organizations.64