Hormone Contraceptives Controversies and Clarifications
Authored by four Christian ProLife Obstetrician-Gynecologists
Recently, there has been some controversy, and serious questions have been raised by sincere individuals who are concerned that hormone contraceptives may have an abortifacient mechanism of action. This paper will help to clarify the issue based on a through review of the available medical literature regarding the mechanism of action of hormone contraceptives. It has been compiled by Board Certified practicing Ob/Gyns, in consultation with Perinatologists and Reproductive Endocrinologists, each being a physician committed to honoring the sanctity of human life from conception. We affirm that as physicians answerable to our Creator and Redeemer, we are responsible to the best of our ability to help, and not intentionally harm, our fellow human creatures. As Christian physicians, we affirm that all life is created by God and that human life is initiated at conception. Fertilization, not implantation, marks the beginning of human life. Disruption of the fertilized egg represents abortion.
The issues of mechanism of action of commonly used hormone contraceptives has threatened to split the pro-life physician community. Review of information currently being disseminated reveals some powerful and well written rhetoric. However, the issue of mechanism of action of hormone contraceptives is not one which will be illuminated by rhetoric. The mechanism of action of any medicine will not change based on how we feel about it, or on who developed it, or on how eloquently it is defended or opposed. How a medication works is a scientific question.
The hormone contraceptives include four basic types: combination oral contraceptives (COCs), injectables (Depoprovera), progestin only pills (minipill, or POPs), and implants (Norplant). In this paper, they will, where convenient, be collectively referred to as the “pill.” Most hormone contraceptives are noted to work by 3 methods of action:
1)Primarily, they inhibit ovulation by suppression of the pituitary/ovarian axis, mediated through suppression of gonadotrophin releasing hormone from the hypothalamus.
2)Secondarily, they inhibit transport of sperm through the cervix by thickening the cervical mucous.
3)They cause changes in the uterine lining (endometrium) which have historically been assumed to decrease the possibility of implantation, should fertilization occur. This presumption is commonly known as the “hostile endometrium” theory.
A thorough review of the medical literature uncovers ample data to support the first two methods of action, which are contraceptive actions. (Appropriate references will be found in the sections discussing each type of hormone contraceptive.) However, there is no direct evidence in the literature to support the third proposed method of action. This conclusion is shared by the respected Gynecologic Endocrinology textbook authors Yen and Jaffe. (1) Nevertheless, for the past nearly 40 years, most authors of “pill” related scientific literature have routinely repeated the assumption of a contra-implantation effect by this “pill” primed uterine lining. In light of this large body of literature, some prolife authors have expressed appropriate concerns that hormone contraceptive methods may include an abortifacient action by hindering implantation. These authors have cited data drawn from this scientific literature to support their claims. Closer scrutiny of the medical literature, however, reveals that the scientists are all simply agreeing that the “pill” produces a thinner, less glandular, less vascular lining. We also agree. However, in an ovulatory pill cycle, the estrogen and progesterone levels are, as discussed below, grossly increased for the seven days prior to implantation. The normal biologic response of endometrium to high levels of these hormones is growth of stroma, blood vessels, glands, and glandular secretions to help prepare the lining for implantation.
An extensive review of pertinent scientific writings indicates that there is no credible evidence to validate a mechanism of pre-implantation abortion as a part of the action of hormone contraceptives. On the contrary, the existing evidence indicates that “on pill” conceptions are handled by the reproductive system with the same results seen with “off pill” conceptions, with the exception of increased ectopic rates seen with POPs and Norplant. Not all the contraceptive agents are equally effective, or even equally appropriate, to be used by doctors and patients concerned with the sanctity of life and maternal welfare. The remainder of this paper is a presentation of the current scientific data which will allow doctors and patients who are committed to the sanctity of life from the time of conception to make decisions regarding the use of these agents that his or her conscience can be at peace with. We do not assume that everyone, given the same information, will arrive at a uniform decision. However, for the follower of Christ, discernment based on prayer and the evaluation of factual information in the light of Scripture is the basis of ethical decision making.
It is helpful at this point to review the basic physiology of the normal ovulatory cycle. Specific endocrinologic details are best found in a text of gynecologic endocrinology. However, in general, after a young woman completes puberty, the levels of estrogen rise and fall twice during each normal menstrual cycle. The pituitary gland releases follicle stimulating hormone (FSH), which causes new, ovum-containing follicles (eggs) to develop in the ovaries during the first half (or follicular phase) of the menstrual cycle. The follicle steadily increases estrogen production, which reaches a peak about one day prior to ovulation. The surge of estrogen stimulates her pituitary gland to secrete another essential hormone, luteinizing hormone (LH), which in turn serves to trigger ovulation (egg release).
Ultrasound can be used to assess the growth and development of the ovarian follicle (cyst around the egg cell) and can indicate the degree of readiness for ovulation.(2) During an ovulatory cycle the usual cyst size varies from 20 to 28 mm. Non-ovulating follicles rarely exceed 14 mm in diameter. Ovulation is associated with complete emptying of the follicular contents in 1 to 45 minutes. After ovulation, the follicle which has released the egg becomes filled with another type of cell, a luteal cell. The luteal cells proliferate under the influence of pituitary luteinizing hormone, (LH), and secrete ever increasing quantities of both estrogen and progesterone.
The follicle (now a corpus luteum) most commonly appears as a smaller, irregular cyst which, if conception has NOT occurred, diminishes in size and ceases to function 2 weeks after ovulation. With subsequent decrease of luteal estrogen and progesterone, the uterine lining (endometrium) is then shed as the menstrual period. However, if conception HAS occurred, the embryo begins, by the time it implants, to secrete another chemical messenger, hCG (human chorionic gonadotropin), which acts like LH to rejuvenate and stimulate the corpus luteum to continue its function until the placenta takes over hormone production 2 months later. The corpus luteum produces, in the six days after ovulation, 10 to 20 times the levels of both estrogen and progesterone seen in a non-ovulatory “pill” cycle. (Preovulatory pill cycle has estradiol level of 25 pg/ml, preovulatory normal cycle has estradiol level of about 40 pg/ml.) During an ovulatory cycle, estradiol reaches a peak of 400 pg/ml during the day before ovulation-a ten to 16 fold increase-and peaks again at 275 pg/ml by day 6 after ovulation, which is the day of implantation. Progesterone values rise from a preovulatory 0.5 ng/ml to a peak of 10 ng/ml by implantation day-a twenty fold rise. (41,42) These high levels act on the lining in theseseven days to prepare it for implantation and support of the arriving (via the fallopian tube) living embryo. Corpus luteum function continues until 8 to 10 weeks from ovulation, at which time (noted above) the placenta assumes the burden of producing these hormones to support the growing pregnancy.
In the extensive literature we have reviewed, no writer has addressed this very significant question:In a menstrual cycle on the “pill” in which ovulation occurs, what is the histology of the endometrium six days after ovulation (the time of implantation)? Certainly the hormone milieu and endometrial histology will be different from a menstrual cycle on the “pill” in which ovulation does not occur(i.e.,the typical atrophic, or “hostile,” endometrium). The FSH-LH-estradiol surge the day before ovulation, and the resulting corpus luteum formation, with its ten to twentyfold estradiol and progesterone output, should produce significant changes in the endometrium. In a normal menstrual cycle, on the day of ovulation the uterine lining (proliferative endometrium) is not receptive to implantation. Seven days of follicle and corpus luteum hormone output transform it to “receptive.” The same follicle and corpus luteum hormone output, when ovulation occurs in a “pill” cycle, should have a similar salutary effect on the pill-primed endometrium. It is highly probable that the so-called “hostile to implantation” endometrium– heralded (without proof) from the beginning by the “pill” producing companies, echoed (without investigation) by 2 generations of scientific writers, and now adopted (as a scientific fact) by some sincere prolife advocates– simply does not exist six days after ovulation in a pill cycle. What is currently known about the endometrial response to corpus luteum hormones suggests this conclusion. Research regarding endometrial histology on the sixth day after escape ovulation in “on pill” cycles would add useful information to the current discussion.
Zanatu (51) reports on two women with prolonged infertility (8 to 14 months) after Depo-Provera injections: “We successfully induced ovulation with the sequential adminstration of clomiphene citrate and human chorionic gonadotropin, and pregnancy immediately followed.” This suggests that once ovulation has occurred, the burst of natural estrogen and progesterone from the corpus luteum simply override even the most potent hormone contraceptive, producing a receptive endometrium, and resulting in a normal implantation and ongoing pregnancy.
The abortifacient theory proponents propose a second line of evidence that they feel strongly suggests the “pill” is associated with an early abortifacient effect. (3) This refers to an increased risk per pregnancy of tubal pregnancy. The lack of a corresponding increase in intrauterine pregnancy is suggested as evidence of a contraimplantation effect of the “pill.” One writer states that “All published data show that the extrauterine ratio of pregnancies is increased for women on BCPs . . .” Our own review of the literature has shown this increased ectopic rate to be true of progestin only pills (POPs) and Norplant. However, we have found absolutely no data in the literature that supports an increased ectopic to intrauterine pregnancy ratio for women using combined oral contraceptives (COCs) or Depoprovera. Comments and references accompany our discussion of these individual agents below.
Two additional lines of reasoning have more recently been offered by abortifacient theory proponents.(3, slides 42-48) The first has to do with integrins, an endometrial polypeptide which is felt by some to be associated with endometrial receptivity. These integrins are “conspicuously absent in patients with luteal phase deficiency, endometriosis, and unexplained infertility….In most OC users, the normal patterns of expression of integrins is grossly altered.” This is felt by the proponent to be evidence of potential abortifacient action at the endometrial level. The problem with this theory is that it deals with endometrium in pill cycles that are not ovulatory. As noted previously, an entirely different hormone milieu exists for seven days to prepare the endometrium for implantation in an “on pill” ovulatory cycle, just as it does in a normal (or “non-pill”) ovulatory cycle. We are aware of no studies dealing with integrins in an ovulatory pill cycle.
The second line of reasoning has to do with endometrial thickness in a pill cycle. (3, slides 35-41) This position notes from the medical literature that “Recent MRI studies show that pill users have endometrial linings that are 40-60% thinner than women not on the pill,” and, “ten recent IVF studies confirm that “endometrial thickness is related to the functional receptivity of the endometrium.” We do not dispute these quotations from the literature. However, as is the case in the previous paragraph, they simply do not apply to endometrial thickness or receptivity in an ovulatory pill cycle, nor do they purport so to apply.
Definition of Abortifacient
Abortion, when used as a medical term, refers to the loss of a pregnancy less than 20 weeks gestational age, regardless of whether the termination is intentional or spontaneous. Spontaneous abortions are commonly known as miscarriages. There are literally hundreds of factors that have been implicated in the loss of human pregnancies. For example, aging of the woman, alcohol, infections, RU486, cocaine, genetic disorders, uterine structural anomalies, methotrexate, some prostaglandins and trauma have all been shown to contribute to abortions. There are many more factors that may contribute to fetal loss, but have not been proven to do so. Implicated factors include almost any environmental substance known to man, including impurities in our air or water, the electronic equipment we sit near, along with almost any medication, herb, or chemical. Some of these may actually cause abortion and some may not.
An ethical question is thus raised, “Where do we draw the line in informed consent for responsible disclosure of known risks?” To answer this question as it relates to hormone contraceptives, we must examine what factors would lead us to classify a drug as an abortifacient.
In order to classify COCs as an abortifacient, several things must exist:
1) Conceptions must occur.
2) The abortive effect must be present with proper use as prescribed.
3) Loss of these conceptions must exceed the base-line loss for populations not using this substance, or be shown to occur due solely to the medication itself, and not other known factors.
4) The abortive effect should be consistent and reproducible by multiple independent observers.
There are no studies that we are aware of regarding combined oral contraceptives or Depoprovera that demonstrate numbers 2,3, or 4. (Increased ectopic pregnancies seen with progestin only pills (POPs) and Norplant will be discussed in the appropriate section below.)
Decision making on the part of a physician prescribing a medicine assumes that the patient will be compliant with the prescribing orders. In our discussion on hormonal contraceptives, patient compliance will be assumed.
Most abortifacient theory proponents have lumped together all of the hormonal contraceptive agents: combined oral contraceptives (COCs), progestin only pills (POPs), implants (Norplant), injectables (Depoprovera). The consideration of rates of ovulation, ongoing pregnancy, and ectopic pregnancy, and the assumption of abortifacient action, has been attributed to all alike. This leads to erroneous conclusions which impact ethical decision making. It is more instructive and accurate to review the literature concerning these agents separately, since they vary in action, complications, and effectiveness.
A careful, exhaustive review of the medical literature has been done. This reveals the following information regarding the mechanisms of action of the four different types of hormone contraceptives:
Progestin Only Pills (POPs)
Progestin only pills have been widely studied, particularly by WHO (World Health Organization) sponsored groups for use in underdeveloped countries. The attractive qualities in this context include probable enhancement of breastfeeding (COCs have negative enhancement), and simplicity of use (one daily, no confusing “dummy” pills or varied colored pills for illiterate populations). In addition, in our cultural context, a high degree of patient tolerance (minimal side effects of nausea, bloating, etc.), no increase in risk (especially in smokers) of estrogen associated thromboembolism, coronary artery disease, or cerebrovascular disease are seen as indications for POP use.(4) However, estimated ovulation rates range from 14 to 84%, (5), varying with age, lactation, and compliance factors. With “typical” use, the average estimated ovulation rate is about 50% (6), ongoing pregnancy rates are estimated at 5% per year. (7) . Pregnancy rates in lactating women, and women over 35 years of age are about l%. (4). Notably, with correct, consistent use, the pregnancy rate for all users is only 0.5% in the first year. (8) To attain this kind of contraception rate, compliance must be unusually exacting (6) The ectopic pregnancy rate with POPs may be between 6 to 10% of unintended pregnancies (6). (In the general population, ectopic pregnancies constitute 2% of all pregnancies.) (9) The increased “per pregnancy risk” for ectopic pregnancy may be a result of progestin action on the tube, reducing the number of cilia on the tubal epithelium, as well as the intensity and frequency of cilia action, thus slowing the rate of ovum transport (10). Others do not confirm this progestin effect on the tube. ( 77) Because of the significant ovulation rate and pregnancy rate, and the increased ectopic pregnancy/ intrauterine pregnancy ratio, the POP may not be a desirable choice for most patients. Elective use of a medication which increases the percent of ectopic pregnancies per conception raises significant ethical questions for many–as does subjecting one’s patient to an increased risk of ectopic pregnancy morbidity or even mortality,when there are equally, or even more effective, alternatives. (In a third world context, additional factors may enter into one’s decision making. See Appendix 1.)
Progestin Implants (Norplant)
The implant Norplant shares many similarities with the POP, except that it is considerably more effective birth control than POPs. Compliance is not a factor.with the implants. The pregnancy rate with Norplant (0.09% per year) is much lower than the pregnancy rate with POPs (0.5% per year with consistent, correct use, 5% per year with “typical” compliance).(8) Therefore the ectopic pregnancy rate per l00 women years of use for Norplant users is also much lower than the rate per hundred women years of use for POP users. In the third world context, Norplant would be about 50 times more effective than POPs at preventing ectopics and maternal mortality (often with attendant newborn mortality), and would facilitate improvement in women’s health and newborn nutrition (see appendix 1). In our cultural setting,, the parameters for ethical decision making are different: High maternal mortality rates are not present, infant nutrition does not depend on continuance of a breast milk supply. Like the POPs, Norplant has an increased EP/IUP ratio, estimated at 3 to 5 times the per pregnancy ectopic rate seen in the general non-contracepted population. The general incidence of ectopic pregnancy in the USA is about 2% of reported pregnancies . Between 6 to 10% of unintended pregnancies on Norplant and POPs may be ectopic. This increased ectopic effect has not been well explained. As mentioned above, it may be related to selective progestin effect on the tube. We see no data to prove or indicate that a number of embryos are produced, with some becoming ectopic pregnancies and with the remainder moving into the uterus to be aborted by an inhospitable lining. But just as in the POP discussion, elective use of a medication with this untoward risk factor, especially when there are hormonal contraceptives with no increased risk factor for either the unborn or the pregnant woman, raises serious ethical questions for many of us.
Injectable DepoProvera (Depomedroxyprogesterone Acetate)
The injectable DepoProvera (DMPA) has only been FDA approved for use in the USA since l992. It has been used internationally since the 1960’s. Although there is a relative paucity of American medical literature pertinent to the topic, there is considerable world experience with DMPA. The literature regarding this injectable suggests its effectiveness is based on extremely low ovulation rates.
The suppressive action of DMPA is at the hypothalamus or higher to prevent the hypothalamus from giving the signal to the pituitary to release gonadotropins.(22) Secretion of LH and FSH are maintained in the mid-follicular state. Because of this, the ovary does not develop a dominant follicle, so the egg does not mature. There is no LH surge, so there is no ovulation. Ovulation does not occur until serum MPA is at extremely low levels, often 7 to 9 months after injection.(20) By the time ovulation is able to occur, serum MPA is at such low levels (O.1 ng/ml) that it has little effect on the endometrium, and the ovary is producing normal preovulatory levels of estradiol. (17, 23 43, 12, 13, 14, 15, 16, 18, 19, 21, 22, 25, 44, 45, 46, 47, 48, 49, 50,) Thus there is no evidence that DMPA produces even a theoretical risk of abortion by “hostile endometrium”.
Several small pharmacokinetic studies utilizing progesterone levels as a major indicator of ovulation have shown a zero ovulation rate 3 months after IM injection of 150 mg of DMPA. (12,13,14,15,16,17,18,19,20) In the practical world, especially the third world, for various reasons the ovulation rate will not be zero (failure to shake the vial, out-dated or deteriorated meds, sub Q rather than IM injection, inadequate dose, biological patient variation all may enter the picture). However, pregnancy rates indicate DMPA is extremely effective. Five large international studies, including over 8,000 women,(21) were used to determine the “Pearl Index” figure of 0.3 pregnancies per hundred women years noted in the chart of “lowest expected pregnancy rates” found in the PDR. The chart is adapted from Hatcher, et al. (8).
There is no evidence that DMPA causes an increased risk of ectopic pregnancy. As noted above, ovulation does not resume until serum progestin levels are extremely low. It is highly unlikely that such negligible progestin serum levels interfere with tubal or tubal cilia motility. In fact, pregnancy is so uncommon with DMPA, that statistics on ectopic pregnancy are difficult to find, and because of very small numbers, difficult to evaluate. The single study identifiable in the literature came from the Downstate Medical Center in New York from l970 to l974. The study looked at gross ectopic rate by reported method of contraception.This study calculated the the gross ectopic rate for DepoProvera of l.3%. However, because the total number of ectopics on DepoProvera was so small, and the total number of pregnancies with DepoProvera was also small, the difference between 1.3% and 1% for the baseline ectopic rate of a comparison group in was not statistically significant. Tatum (24), commenting on the study, notes, “Since pregnancy on depo-provera was so rare, there were not enough pregnancies to even get a statistically accurate rate.” The significance of these studies is that they show no dramatic increase in ectopic pregnancies. Rather, if pregnancy does occur with DMPA, it is about as likely to be ectopic as a non-contracepted pregnancy.
Combined Oral Contraceptives (COCs)
Over the last 30 years, combined oral contraceptives (COCs) have become one of the most commonly used forms of family planning world-wide. Billions of women have benefited from both the direct contraceptive and indirect health-related benefits of these medications. Ninety nine per cent of American women using hormone contraception use COCs. Although there are risks associated with use of any medication, the risks of COCs are extremely small, especially when patients are appropriately chosen and counseled for use.
Definition of Combined oral contraceptive (COC)
Combined oral contraceptives (COCs) are those birth control pills that contain both an estrogenic and a progestational agent. There are several different estrogens and even more progestins available on the market, and there are many ways to combine them into pill formulations. The contraceptive efficacy of these hormones is derived primarily from the actions of progestins in suppressing ovulation and thickening cervical mucus. The estrogenic component serves two purposes. It potentiates the action of the progestogenic component of oral contraceptives, and it stabilizes the endometrium so that irregular shedding and unwanted breakthrough bleeding can be minimized. (25).
In fact, an ideal endometrium does not seem to be necessary for successful implantation, as is evidenced by ectopic pregnancies which may implant in the tubes, endocervical canal, on the ovaries, or on the surface of any intra-abdominal organ, including bowel or even major blood vessels. Additionally, as noted above, in an “on pill” escape ovulatory cycle, with the required FSH-LH surge, followed by post ovulatory corpus luteum estradiol and progesterone output, one would expect the endometrium to undergo the usual hormonally related changes favorable to implantation, as in any ovulatory cycle. (The endometrium is sufficiently responsive to physiologically balanced hormones that even the slightly increased estrogen balance in triphasics produces a histologic trend toward secretory pattern.)(40) To have a meaningful discussion regarding the mechanism of action of COCs, and to address the ” potential abortifacient” question, a review of the pertinent medical literature is necessary. The following discussion is based on our review.
Regarding the Occurrence of Conceptions on COCs with Proper Use
Conceptions do occur, at times, on COCs. We know this, because like all methods of contraception other than abstinence, there is a measurable pregnancy rate. Medical literature often refers to pregnancy rates for contraceptive methods as “failure rates.” So called failure rates for contraceptive methods are classified as perfect use (compliant, consistent) or typical use. Perfect use failure rates imply that the method of contraception has failed, whereas typical use values reflect human noncompliance or other confounding variables (such as pharmaceutical interactions) that contribute to failure. Since this ethical discussion regarding COCs is primarily about the method of action of COCs, we are interested in perfect use failure rates, which most accurately reflect the activity of the drug itself. The perfect use failure rate most often quoted in the medical literature, including the standardized FDA product labeling of every combined oral contraceptive listed in the PDR is 0.1 pregnancies for every 100 women years (1300 cycles). Therefore, it can be concluded that at times, ALL of mechanisms of action of COCs fail.
Regarding spontaneous abortion rates with “pill” pregnancies.
Spontaneous abortions can be divided functionally into “clinical abortions,” i.e., spontaneous loss after pregnancy has been clinically recognized (usually from about 6 menstrual weeks, which is 4 conceptional weeks), and “pre-clinical abortions,” loss occurring before that time. Pre-clinical abortions can further be divided into “pre-implantation abortions, those occurring before the conceptus implants 6 days after fertilization, and those occurring after implantation, but before clinical recognition of the pregnancy. There has been no demonstrated effect of COC use on spontaneous clinical abortion rates. (78) The essential abortifacient argument brought against hormonal contraception is that it causes pre-clinical, more specifically, pre-implantation, abortion due to an inhospitable lining, the”hostile” endometrium.” The improbability of this entity, based on known follicle and corpus luteum hormone output during an ovulatory cycle, and normal endometrial response to these hormones, has been discussed earlier in this paper.
To look at the controversy from another direction, do “pill” pregnancies have similar outcomes as non-pill pregnancies? Clinically, the answer is “yes.” There is no data to indicate higher clinical spontaneous abortion rates or more problems in ongoing pregnancies. But is there increased loss evident with “pill” pregnancies? From the clinical side, the answer is “no.” (78) From the pre-clinical, especially the pre-implantation perspective, the answer must be sought by evaluating indirect data, since there is no direct data regarding these loss rates available for users of COCs. Most studies evaluating efficacy of COCs only measure clinically evident pregnancies as an end point. There is scarce literature about ovulation rates on COCs, although more than 40 such studies were reviewed in preparing this paper. Of the COC studies that evaluated ovulation, fertilization and pregnancy rates are almost never evaluated. The reason for this should be obvious: if a patient in a COC study is told that she has ovulated, she will avoid exposure to sperm, thus preventing an unwanted pregnancy.
Concerning the outcome of “on pill” ovulations
The first requisite in evaluating this question is to establish a reliable unintended ovulation rate in perfect (compliant, i.e., no missed pills) COC users. In this pursuit, 25 studies were reviewed (several papers contained more than one study). (ref 59 through 76) These studies used a variety of common COCs, including triphasics, and were about evenly split between the newer very low dose pills(20 mcg estinyl estradiol), and the current standard low dose pills(30 to 35 mcg ethinyl estradiol). Eighteen studies including 3799 cycles showed zero ovulations. Seven studies including 2910 cycles showed 8 ovulations. (Ovulation was indicated by ultrasound and serum chemical markers.) Combining these gives a practical working number of 8 ovulations in 6709 cycles, which equals 15.5 ovulations in 13,000 cycles, (a figure used to simplify the arithmetic done below). This is not a scientific metanalysis. Rather, it is a pragmatic figure, using the referenced peer-reviewed data, that will help provide an informed perspective on the question of pre-implantation loss.
The next necessary information concerns the unintended pregnancy rate on compliant use of COCs. This is well established in the Hatcher table (8) at 0.1 pregnancy per 100 women years, which equals one pregnancy per 1,000 women years, or one pregnancy in 13,000 cycles by compliant pill takers. Thus we have, in l3,000 cycles, l5.5 ovulations (from the previous paragraph), and one pregnancy.
Finally, the cervical mucus factor must be considered. The marked change in cervical mucus under the influence of progestins is recognized as a substantial factor in contraceptive effectiveness of the “pill.” As reviewed by McCann, (6) studies of cervical mucus changes on the POPs, which contains only half the dose of progestin found in the COCs, found the mid cycle mucus to be greatly reduced in volume, increased in viscosity and cell content, with altered molecular structure. (52, 53, 54) The effect is a mucus with low spinnbarkeit and poor ferning. This is found to result in little or no sperm penetration in 70-80% of cases. (53) Even in the rare cases when penetration does occur, sperm motility is reduced.(55, 54, 56, 57, 58) One study noted almost total absence of sperm in the uterine cavity of the progestin treated group, while sperm were present in the uterus of 18 of 19 controls. (55) COCs, with twice the progestin dose of POPs, would produce mucus with similar, if not greater sperm impedance. (Although COCs also contain an estrogen, ethinyl estradiol, they block follicle activity so well that the actual preovulatory serum estradiol levels on the “pill” are less than normal, i.e.,25 pg/ml vs the normal 40 pg/ml. This level will have negligible influence on improving cervical mucus.) In the normal cycle, the pre-ovulatory FSH surge will immediately produce an estradiol peak of 400 pg/ml, quickly resulting in production of the optimum fertility enhancing mucus. In an ovulatory “pill” cycle, this estradiol peak almost certainly will override, to some degree, the progestin induced sperm blocking effect on the cervical mucus. It is not known how quickly, or to what degree, this override might take place. Reproductive endocrinologists indicate that ovulation takes place within 12 to 24 hours after the LH-FSH surge with its accompanying estradiol peak, and that the newly released egg can only accept fertilization for about l2 hours before it becomes resistant. This leaves a window of about 24 to 36 hours for the transformation of the impenetrable cervical mucus to fertility-favorable mucus which would allow release and transport of sperm to the distal portion of the fallopian tube to fertilize the egg. (And this assumes that sufficient numbers of viable sperm are present in the cervical mucus at the time the mucus becomes favorable.) We know that the cervical mucus factor adds significantly to the contraceptive effectiveness of the pill–although to what extent in an ovulatory cycle can only be estimated, since there have been no specific studies done to give us numbers.
The available data referenced in this discussion has, then, yielded this information: 13,000 cycles of compliant COC use results in 15.5 ovulations, and one ongoing pregnancy. The basic question in the entire “pill” controversy is, “What happened to the other 14.5 eggs?” Were they the victim of an “inhospitable endometrium”(whose existence in an ovulatory cycle is very questionable), or can they be accounted for in other ways? Let us start with the 15.5 ovulations noted above. Available data indicate that 10 to 15% of our population is infertile.(41, p 809) Of the original 15.5 eggs, This leaves 13.5 eggs available for possible fertilization and ongoing pregnancy. One could simply take the 13.5 eggs available for fertilization, apply the normal human fecundity table of 25% per cycle (the fecundity table requires optimum conditions: normal fertility in the man and woman, and correct timing of coitus). (28) We would expect 3.3 pregnancies, IF the cervical mucus is favorable, the viable sperm are present, and the timing is right. Referencing the previous paragraph, cervical mucus factors block sperm 80 to 90% of the time in non-ovulatory studies. IF adequate numbers of viable sperm can get free and make it to the waiting egg within the 36 hour receptivity window just 33% of the time, one egg might be fertilized to become an ongoing pregnancy. We don’t know about the sperm and timing, and the cervical mucus is probably not optimum. If one of these 3.3 remaining eggs is fertilized and thrives, this equals the known ongoing pregnancy rate seen in compliant COC users: One per 1,000 woman years (13,000 cycles) of use. The fate of the l5.5 eggs can also be considered from other available data. Again, the 10 to l5% infertility rate leaves 13.5 eggs available for possible fertilization. Recent critically reviewed data indicates that from fertilization to 6 weeks, spontaneous pregnancy wastage is 73%. (26, 27) This leaves 3.5 eggs available for fertilization and ongoing pregnancy. As above, considering mucus, sperm, and timing factors, one egg might be fertilized to become an ongoing pregnancy. This again equals the known pregnancy rate on compliant COC usage: per 1,000 woman years (13,000 cycles) of use. This exercise in arithmetic is not meant to be a statistician’s scientific proof text. It is rather an overview of COC unintended ovulation rates and subsequent ongoing pregnancy rates using available peer-reviewed data to account for the eggs in question.
A recent study by specialists in reproductive technology found, with genetic sampling of morphologically normal embryos, that 24 of 50 had chromosomal abnormalities, and would likely not survive. (79) Gift procedures, introducing multiple viable eggs, along with good sperm, into the fallopian tube at the optimum time, with optimum endometrium, only yield a 30% success rate. 30% of failed fertilization is due to faulty sperm. There are evidently many naturally occurring reasons for preimplantation loss.
Considering the above information on ovulatory related hormone influences on the endometrium, “on pill” ovulation rates and pregnancy rates, the cervical mucus factor, and known data on human fecundity, fertility, and spontaneous loss rates, it can be appreciated that there is no need to postulate a pill-induced preimplantation abortion phenomenon to explain why 15.5 eggs produce one ongoing pregnancy. Known and natural causes can account for the numbers.
Concerning Ectopic Pregnancy and the “Pill”
If COCs are abortifacient in nature by causing the conception not to implant in the uterine cavity, then ectopic pregnancy rates should be at least equal to that in the normal population. One of the well-studied effects of progestins on the fallopian tube is to decrease motility and cilia action, which would inhibit a fertilized ovum from proceeding into the uterine cavity. Therefore, it would be reasonable to expect an even higher ectopic pregnancy rate than the normal population (2% of all clinically recognized pregnancies). The literature, at this time, does not show an increased ectopic rate per pregnancy for COCs over non-users.
Any contraceptive method decreases the overall ectopic rate, simply by way of decreasing the number of pregnancies that occur at all. A rough estimate of the numbers involved can be calculated using generally accepted data and simple arithmetic: Using the table of Hatcher (8), and assuming “perfect” (correct, consistent) use for COCs, sympto-thermal method, and condoms, three matched groups of l00,000 women would experience l00 unintended pregnancies on COCs, 2,000 unintended pregnancies with sympto-thermal method, and 3,000 unintended pregnancies with condoms. Using the current figure of 2% ectopic ratio (9), these pregnant women would experience 2 ectopic pregnancies with COCs, 40 EP with sympto-thermal method, and 60 EP with condoms. Given the potential maternal mortality, morbidity, and damage to fertility with ectopic pregnancy, and the absence of any demonstrated increase in abortion due to COC endometrial effect, one might ask if it is ethical to withhold or discourage the option of COCs for the woman seeking contraceptive information. Certainly this pertinent information should be considered as part of adequate informed consent.
Although there are many references alluding to increased rates of ectopic pregnancies for tubal ligation, IUDs, POPs, and post-coital contraception, a search through over a dozen well-recognized medical texts, and multiple journal articles written by experts in obstetrics, gynecology, and contraception has yielded no authoritative opinions that implicate COCs in the etiology of ectopic pregnancy. Additionally, medical society as a whole has had over 40 years of experience with COCs, with virtually millions of women world wide, and a history of billions of cycles of use. In all of this time, although many other risks of COCs have been identified (and are part of the standardized labeling of all COCs in the PDR, as required by the FDA), not one pharmaceutical company, independent clinical researcher, or epidemiologist has published data clearly showing an association between COCs and increased ectopic pregnancy rates. In the medico-legal world, product liability implications have an immense influence on standardized labeling. If, in the 40 years of COC experience, there were the slightest hint that the product might cause an increase in ectopic pregnancy rates, such warning would have been mandated in the product literature to protect the manufacturer from lawsuits, as well as to warn the user of the possible complication. There are no such warnings in COC labeling. An ectopic pregnancy noted in a woman who has compliantly used COCs, and has no other major risk factors for ectopic pregnancy, is so rare that it would be a reportable case.
A careful search of the literature reveals 11 references to ectopic pregnancies in COC users (29-39). Of those 11, one paper (32) calculates the theoretical risk of ectopic pregnancies in COC users, the other 10 are clinical papers. Of these 10 references, one (34) is a letter to the editor, which references data from a clinical study (29). References 29 and 31 are written by the identical group of researchers, with identical data, just published in two different journals, in two different languages.
In particular, one reference (35) is often quoted by abortifacient theory proponents as suggesting a “slightly” increased ectopic pregnancy rate on COC’s. A critical review of this article reveals this to be a misleading statement. This meta-analysis attempts to draw conclusions from the small amount of data from references 33, 36, 37, and 38. The actual data review compiles data from these 4 studies into 3 categories: 1) one shows a strong, statistically significant decrease in ectopic pregnancies in COC users, even supporting a protective effect against ectopic pregnancy. 2) the two other groups analyzed showed no statistically significant increase in ectopic rates among COC users, even when these patients were compared to inappropriate control groups of pregnant patients. Therefore, the author’s conclusion that suggested a possible increased rate of ectopics with COC use is not supported either in their own meta-analysis data, or by any of the individual papers used in the analysis.
One reference (39) is an epidemiologic study from Beijing, China. This is the only study that gives an ectopic pregnancy rate in women on COCs. The data is expressed in the form of ectopic pregnancies per 1000 woman cycles. Unfortunately, there is not enough information in the paper to make a comparative statement about whether this rate is higher or lower than the normal population in Beijing, since ectopic rates are generally reported as a percentage of clinically recognized pregnancies, not cycles. Also, these authors provided no data concerning other risk factors that these women may have had for ectopic pregnancies. In fact, the risk may have been considerable, since the group studied were “sex workers.”
Of the remaining 6 references with clinical data, none were designed to evaluate ectopic rates in COC users. In fact, most of these studies use COC users as their control group in evaluating other contraceptive methods. None of these studies calculated ectopic rates per clinical pregnancy for COC users, and all of them describe their data in the form of the number of women who happened to be taking COCs when they conceived an ectopic pregnancy. Several of these studies showed ectopic rates for barrier methods of contraception (which have no impact at all on the endometrium or implantation) to be higher than for COCs. None of these authors conclude that they have data proving ectopic rates to be higher in COC users when compared to non-users. One author (29) makes a suggestion based on 11 patients with ectopics who were on COCs that COCs may prevent intrauterine implantation better than tubal implantation. However, this study has small numbers, too small to reach statistical significance, and uses term pregnant women as the comparison group. This is acknowledged by the authors to be a poor choice of a control group, for the reason that they have selected out any women who had spontaneous abortions, early intentional terminations, or ectopic pregnancies. Furthermore, even though this author makes this suggestion, he makes it clear that it is not conclusive evidence, as the formulations of the pills and patient compliance were not assessed, and he suggests that future research is needed to verify his results.
An analytical error which must be avoided in reviewing the above literature is attempting to calculate an ectopic pregnancy rate based on the numbers in the studies above. This is faulty logic: calculating the number of ectopic pregnancies that occur in women taking COCs is not the same as the ectopic pregnancy rate (i.e.,the fraction of women with ectopic pregnancies among those who became pregnant as a result of COC failure.) The sampling frames are different.
None of the studies listed above even estimate the sample size of pill users that are represented by the few ectopic pregnancies reported. None of the studies specifically evaluated patient compliance or the multiple other risk factors that women in the ectopic groups may have had in addition to oral contraceptive use, and most studies do not separate out progestin-only preparations from combined oral contraceptives.
Therefore, at this time, there is no direct evidence that shows that combined oral contraceptives are an independent risk factor for ectopic pregnancy when conception occurs. Given the medical community’s long history of experience with COCs, there is no evidence that COCs are an etiologic risk factor for ectopic pregnancy. Therefore, the existing medical literature does not support the theory (with its “abortifacient” implications) that COCs are more efficacious
in preventing intrauterine pregnancy than ectopic pregnancy.
Given the above, there is no evidence that shows that the endometrial changes produced by COCs contribute to failure of implantation of conceptions, nor is there evidence that COCs cause an increased per pregnancy ratio of ectopics. .
Some Summary Thoughts
l. Ethical decision making by a medical doctor often involves life and death questions. The discussion regarding the ethics of hormonal birth control is an extremely important matter, since l4 million women in the United States alone regularly use these medications (and that is but a fraction of the world wide usage).
2. Ethical decision making on this issue must involve the evaluation of as much pertinent scientific information as possible.
3. Christian ethical decision making requires that the physician be committed to the Authority of Scripture and to the Lordship of Jesus Christ in his/her life and medical practice.
4. Christian ethical decision making on this issue requires the physician be committed to the sanctity of life of the unborn from the time of conception, and to the sanctity of life of the mother.
5. The patient must be a part of this ethical decision making, and must be informed of the pertinent issues involved.
The following questions should be addressed by pro life physicians:
1) Is it appropriate to implicate a medication as an abortive agent without the data to support such a claim? To do so creates needless hostility and division among physicians and patients who genuinely respect life from the moment of conception.
2) Where do we draw the line in informed consent for responsible disclosure of known medical risks vs. a theoretical risk which is not substantiated by current scientific knowledge?
3) Is it accurate to implicate all hormonal contraceptive methods as one regarding their method of action, rather than evaluating each one individually?
We have done our best to evaluate the current literature on the topic, and to present it fairly for consideration by the reader. We do not find substantive evidence that hormone contraceptives include an abortifacient mechanism of action. We recognize that equally spiritual and honest men and women may consider the same facts, but come to differing conclusions on this matter. For the follower of Jesus, the decision on whether or not to use or prescribe hormone contraceptives is a spiritual as well as a medical matter. The principles of how to handle a situation among Christian Believers on a very vital subject in dispute is directly addressed in Romans chapter l4 as an issue of individual conscience and responsibility before the Master (see appendix 2).
Many factors play a part in how a family plans and spaces their children. It is not the purpose of this paper to promote nor to oppose hormone contraception. However, if a family, weighing all the factors affecting their own circumstances, decides to use this modality, we are confident that they are not using an abortifacient.
This paper is not meant to be the “final word” on this issue. If scientific study should validate that a hormone contraceptive agent is partly abortifacient in its action, we would oppose that agent just as we oppose elective medical and surgical abortions. We must constantly examine valid data as it becomes available in our effort to discern what are appropriate methods of family planning to be used or prescribed by those who know that human life begins at the time of conception.
We wish to state emphatically that the following discussion does NOT refer to any program intent on world population control, or what is sometimes called “population control imperialism.” Rather, it refers to compassion and treatment for patients under our care. As Christian physicians, it is our mandate to alleviate, to the extent humanly possible, medical conditions resulting in suffering and death. In 3rd world countries, maternal mortality rates are much higher than in our setting, and the use of hormone contraceptives will mean significantly decreased maternal mortality. For example, 18 countries in Africa have a maternal mortality from 100 to more than 200 times that in the United States.(ll) In the US, the maternal mortality rate is 7 maternal deaths per l00,000 pregnancies. In Zambia, it is 764 to l549 maternal deaths per l00,000 pregnancies. Their hospital based maternal mortality is 543/l00,000, or 80 times the US rate. In this context, POPs, the least effective hormonal contraceptive, would result in an immense decrease in pregnancy related deaths. Using 85 per l00 women years pregnancy rate in non-contracepting women from Hatcher (8), using the “typical” 5% unintended “pill”pregnancy rate from Trussel (7), using the highest (l0% of pregnancies) ectopic rate (6), using the 543 per 100,000 hospital based maternal mortality from northern Zambia, and using a “normal” ectopic rate of 2%, we would see, among l20,000 women using POPs: 5400 pregnancies rather than l00,000 pregnancies, 540 ectopics rather than 2000 ectopics, 29 late pregnancy-delivery related deaths rather than 543 deaths. These are obviously approximate figures, but they come from simple arithmetic with the most accurate data base available. Additional benefits would include improved health for mothers not subject to the drain of a rapid succession of pregnancies, and improved infant nutrition with enhanced and prolonged breast feeding possible before the next newborn displaces its older sibling from the breast (the “kwashiorkor” phenomenon).
In Romans l4, the issue was eating meat offered to idols, a violation of the First Commandment. In the present controversy, the issue is the killing of unborn children, a violation of the Sixth Commandment. We recognize that equally spiritual and honest men and women may consider the same data, but come to differing conclusions on this matter. Thus we face, in the terminology of Romans l4, a “disputable matter” among Believers. The principles of how to approach such a vital issue are laid out for us in Romans l4. We bring attention to these principles without implication that one side or the other in this “disputable matter” is the “weaker brother,” since we are all weak in many ways. Christ is our only strength. Rather, we seek behavior and attitudes amidst this controversy that will “lead to peace and mutual edification” (v 19). First, we are commanded to accept one another “without passing judgment on disputable matters” (v 1). We are instructed to “not look down upon” nor “condemn” one another (v 3), nor to “judge your brother” (v 10). “For who are you to judge someone else’s servant? To his own master he stands of falls. And he will stand, for the Lord is able to make him stand” (v 4). Rather, we are instructed to be “fully convinced in our own minds” (v 5), and to make decisions and perform actions “to the Lord” (vv 6,8). For “each of us will give an account of himself to God” (v 12). It is noteworthy that either of the disputable behaviors, eating or abstaining, may please, or displease, God. God, rather, has other priorities. “For the Kingdom of God is not a matter of eating and drinking, but of righteousness, peace, and joy in the Holy Spirit” (v 17). And finally, in our decision making on this issue, “everything that does not come from faith is sin” (v 23).
This material will be found following the bibliography (below).
Susan A. Crockett, MD Donna Harrison, MD
Joe DeCook, MD Camilla Hersh, MD
1. Yen SSC, Jaffe RB. Contraceptive hormones and their clinical use. In: Reproductive endocrinology: physiology, pathophysiology, and clinical management, 3rd edition. Philadelphia, WB Saunders. 1991. Pp 814-815.
2. Ritchie. Ultrasound in the evaluation of normal and induced ovulation. Fertil Steril . 1985;43:167.
3. Larimore, WL. Is the Birth Control Pill An Abortifacient? WELS Lutherns for Life Bioethics Conference, Orlando, Florida, Feb l8, l999, slides49-61.
4. Kaunitz, AM. Revisiting progestin-only oral contraceptives. Contemporary Ob/Gyn, 1997; Dec:92-98.
5. Landgren BM, Diczfalusy E. Hormonal effects of the 300ug norethindrone (NET) minipill. I. Daily steroid levels in 43 subjects during a pretreatment cycle and during the second month of NET administration. Contraception. 1980;21:87-113.
6. McCann MF, Potter LS. Progestin-only oral contraception: a comprehensive review. Contraception. 1994;50(Supp. 1):21-53.
7. Trussel J. (Princeton University Office of Population 6. McCann MF, Potter LS. Progestin-only oral contraception: a comprehensive review. Contraception. 1994;50(Supp. 1):21-53. Research, Princeton, NJ) Letter to: Margaret McCann and Linda S. Potter. June 19, 1993.
8. Hatcher RA, Trussell J, et al, eds. Contraceptive Technology, 16th rev. ed. New York: Irvington Publishers, Inc, 1994.
9. Morbidity/Mortality Weekly Report from CDC, Vol. 44, No. 3, Jan 27, 1995, p. 46.
10. Li XF, Newton J. Progestogen-only contraceptives and ectopic pregnancy. Br. J. Fam Plann. 1992;18:79-84.
11. Revised 1990 Estimates of Maternal Mortality: A New
12. Bassol, S. et al: Ovarian function following a single administration of depo-medroxy progesterone actate (DMPA at different doses. Fertil. Steril. 1964;42(2):216-222.
13. Fotherby, K et al: Pharmakokinetic study of different doses of depo-provera. Contraception. 1980;22(5):527-36.
14. Fotherby, K. et al: A preliminary pharmakokinetic and pharmacodynamic evaluation of depot-medroxyprogesterone acetate and norethisterone oenanthate. Fertil. Steril. 1980;34(2):131-9.
15. Fotherby, K. et al: Return of fertility in women discontinuing injectable contraceptives. J. Obstet. Gynecol. 1986;6(Suppl 2):5110-5.
16. Garza-Flores, et al: Return to ovulation following the use of long acting injectable contraception. A comparative study. Contraception. 1985;31:361-4.
17. Jeppson S. et al: Endometrial histology and circulating levels of medroxyprogesterone acetate (MPA), estradiol, FSH, and LH in women with MPA induced amenorrhea compared with women with secondary amenorrhea. Act. Obstet. Gynecol. Scand. 1977;56:43-8.
18. Jeppson, S. et al: Medroxyprogesterone acetate, estradiol, FSH and LH in peripheral blood after administration of Depo-provera to women. Contraception. 1976;14(4):461-9.
19. Mishell, D Jr. et al: Estrogenic activity in women receiving an injectable progestogen for contraception. Am J. Obstet. Gynecol. 1972:113(3);372-6.
20. Ortiz, A. et al: Serum medroxy progesterone acetate (MPA) concentrations and ovarian function following intramuscular injection of Depo-MPA. JCE&M 1977;44(1):32-8.
21. Trussell, J. et al: Contraceptive failure in the United States: a critical review of the literature. Stud. Fam. Plan. 1987;18(5):237-83.
22. Toppozada, M. et al: Effect of injectable contraceptives depo-provera and norethisterone oenonthate on pituitary gonadotropin response to luteinizing hormone-releasing hormone. Fertil. Steril. 1978;30(5):545-8.
23. Ludwig, H. The morphologic response of the human endometrium to long term treatment with progestational agents. Am. J. Obstet. Gynecol. 1982;142(6, part 2):796-8.
24. Tatum HJ, et al. Contraceptive and sterilization practices and extrauterine pregnancy: a realistic perspective. Fertil. Steril. 1977;28(4):407-21.
25. Lobo RA, Stanczyk FZ. New knowledge in the physiology of hormonal contraceptives. Am J Obstet Gynecol. 1994;170:1499-507.
26. Boklage CE. Survival probability of human conceptions from fertilization to term. In J Fert. 1990;35(2):75-94.
27. American College of Obstetricians and Gynecologists, Committee on Technical Bulletins. Early Pregnancy Loss. Technical Bulletin 212, 1995.
28. Cunningham FG, McDonald PC, et al. Abortion, Chapter 26, pp. 579-82. In: Williams Obstetrics, 20th ed. Stamford, Connecticut, Appleton & Lange. 1997.
29. Coste J, Job-Spira N, et al. Risk factors for ectopic pregnancy: a case-control study in France, with special focus on infectious factors. Am J Epidemiol. 1991;133:839-49.
30. De Muylder X. Ectopic pregnancy in Zimbabwe. Int J Gynaecol Obstet. 1991;35(1):55-60.
31. Fernandez H, Coste J, et al. Risk factors of ectopic pregnancy. A case-control study at 7 maternity units in the Paris area. J Gynecol Obstet Biol Reprod. 1991;20(3):373-9.
32. Franks AL, Beral V. Contraception and ectopic pregnancy risk. Am J Obstet Gynecol. 1990;163:1120-3.
33. Holt VL, Chu J, et al. Tubal sterilization and subsequent ectopic pregnancy: a case-control study. JAMA. 1991;266(2):242-6.
34. Job-Spira N, Fernandez H, et al. Letter of response to editorial comment, “Risk of Chlamydial PID and Oral Contraceptives.” JAMA. 1990;264(16):2073.
35. Mol BWJ, Ankum WM, et al. Contraception and the risk of ectopic preganancy: a meta-analysis. Contraception. 1995;52:337-41.
36. Ory HW, and the Women’s Health Study. Ectopic pregnancy and intrauterine contraceptive devices: new perspectives. Obstet Gynecol. 1981;57:137-44.
37. Thorburn J, Berntsson C, et al. Background factors of ectopic pregnancy. I. Frequency distribution in a case-control study. Eur J Obstet Gynecol Reprod Biol. 1986;23:321-31.
38. World Health Organization. A multinational case-control study of ectopic pregnancy. The World Health Organization’s special programme of research, development and research training in human reproduction: task force on intrauterine devices and fertility regulation. Clin Reprod Fertil. 1985;3(2):131-43.
39. Zhang Z, Liju W, et al. An epidemiological study on the relationship of ectopic pregnancy and the use of contraceptives in Beijing – the incidence of ectopic pregnancy in the Beijing area. Contraception. 1994;50:253-62.
40. Robertson WB: Effects of steroids on the endometrium. In New Considerations in Oral Contraception, ed by I Brosens. New York, Biomedical Information Corporation Publications, l982, pp153-171.
4l. Speroff, Glass, and Kase “Clinical Gynecologic Endocrinology and Infertility, 5th edition, Williams & Wilkins, Baltimore, 1994.
42. Brenner PF, Mishell DR,Jr., Stanczyk FZ, et al: Amer J Obstet Gynecol 129:133 1977.
43. Antal, Ellen G Pharm D. BCOP Drug Information Clinical Pharmacist, Pharmacia and Upjohn, Personal correspondence.
44. Froser, I: Long acting injectable hormonal contraceptives. Clinical Reproduction and Fertility l l982 pp67-88.
45. Bodlzieher, W et al: A cross sectional study of plasma, kFSH, and LH levels in women using sequential, combined or injectable steroid contraceptives over long periods of time.
46. Heber, KR: Medroxyprogesterone acetate as an injectable contraceptive. Australia Fam. Phys. l7(3) March l998 pp 199-204.
47. Jeppson, S: Background development and properties of progesterones, with special reference to medroxyprogesterone acetate (MPA). Acta Obstet Gynecol Scand Suppl101,7-10 l981.
48. Kaunitz, A: Long-acting injectable contraception with depot medroxyprogesterone acetate. Am J. Obstet. Gynecol. 170(5) Part 2 1543-1549.
49. Kirtin, K et al: Return of ovulatory cyclicity following an intramuscular injection of medroxyprogesterone acetate (provera). Contrac. 10(1) l974 pp39-45.
50. Schwallie, PC: The effect of depomedroxyprogesterone acetate on pituitary and ovarian function, and the return of fertility following its discontinuation: a review. Contrac. 10(2) August l974 pp181-202.
51. Zanartu, J.: Long term contraceptive effect of injectable
progestogens: Inhibition and reestablishment of fertility Int. J. of
Fertility Oct-Dec 1968 Vol 13 No 4 pp 415-426.
52. Chretien FC, Sureau,C, Neau C. Experimental study of cervical blockage induced by continuous low-dose oral progestogens. Contraception 1980;22:445-56.
53. Martinez-Manautou J, Giner-Velasquez J, Cortes-Gallegos V, Aznar R, Rojas B, Gueterrez-Najar A,et al. Daily progestogen forcontraception: A clinical study. Br. Med J 1967;2:730-2.
54. Moghissi KS, Syner FN, McBride LC. Contraceptive mechanism of micro-dose norethindrone. Obstet Gynecol 1973; 41:585-94.
55. Kesseru-Koos E. Influence of various hormonal contraceptives on sperm migration in vivo. Fertil Steril l971;22:584-603.
56. Roland M. Prevention of sperm migration into the uterine cavity by a microdose progestagen. Fertil Steril 1970;21:211-6.
57. Ruiz-Velasco V, Mariscal JDO, Salgado HJ, Salinas GR, Zelaya AP, PartidaJB. Ethynodiol diacetate as a contraceptive. Fertil Steril 1974;25:927-34.
58. Schumacher GFB. Immunology of spermatozoa and cervical mucus. Hum Reprod 1988;3:289-300.
59. Crosignani PG, Testa G, et al. Ovarian activity during regular oral contraceptive use. Contraception. 1996;54(5):271-3.
60. Csemiczky G, Dieben T, et al. The pharmacoldynamic effects of an oral contraceptive containing 3 mg micronized 17 beta-estradiol and 0.150 mg desogestrel for 21 days, followed by 0.030 mg desogestrel only for 7 days. Contraception. 1996;54(6):333-8.
61. Egarter C, Putz M. Ovarian function during low-dose oral contraceptive use. Contraception. 1995;51(6):329-33.
62. Fitzgerald C, Feichtinger W, et al. A comparison of the effects of two monophasic low dose oral contraceptives on the inhibition of ovulation. Adv Contracept. 1994;10(1):5-18.
63. Fitzgerald C, Feichtinger W, et al. A comparison of the effects of two monophasic low dose oral contraceptives on the inhibition of ovulation. Adv Contracept. 1994;10(1):5-18.
64. Killick SR, Firzgerald C, et al. Ovarian activity in women taking an oral contraceptive containing 20 ug ethinyl estradiol and 150 ug desogestrel: effects of low estrogen doses during the hormone-free interval. Am J Obstet Gynecol. 1998;179(1):S18-24.
65. Lunell NO, Carlstrom K, et al. Ovulation inhibition with a combined oral contraceptive containing 20 ug ethinyl estradiol and 250 ug levonorgestrel. Acta Obstet Gynecol Scand Suppl. 1979;88:17-21.
66. Rossmanith WG, Wirth U, et al. Ovarian activity, cycle behavior and tolerance of low dosage oral contraceptives: a comparative study. Zentralbl Gynakol. 1997;119(11):538-44.
67. Schubert W, Cullberg G. Ovulation inhibition with 17 beta-estradiol cyclo-octyl acetate and desogestrel. Acta Obstet Gynecol Scand. 1987;66:543-7.
68. Spona J, Lachnit-Fixson U, et al. Inhibition of ovulation by a triphasic gestodene-containing oral contraceptive. Adv Contracept. 1993;9(3):187-69. Spona J, Elstein M, et al. Shorter pill-free interval in combined oral contraceptives decreases follicular development. Contraception. 1996;54(2):71-7.
70. Spona J, Feichtinger W, et al. Inhibition of ovulation by an oral contraceptive containing 100 ug levonorgestrel in combination with 20 ug ethinylestradiol. Contraception. 1996;54(5):299-304.
71. Spona J, Feichtinger W, et al. Modulation of ovarian function by an oral contraceptive containing 30 ug ethinyl estradiol in combination with 2.00 mg dienogest. Contraception. 1997;56(3):185-91.
72. Teichmann AT, Brill K, et al. The influence of the dose of ethinylestradiol in oral contraceptives on follicle growth. Gynecol Endocrinol. 1995;9(4):299-305.
73. Thomas KH, Vankrieken L. Inhibition of ovulation by a new low-dose monophasic contraceptive containing gestodene. Int J Fertil. 1989;34(Suppl):10-21.
74. Van der Does J, Exalto N, et al. Ovarian activity suppression by two different low-dose triphasic oral contraceptives. Contraception. 1995;52(6):357-61.
75. Wenzl R, Bennink HC. Ovulation inhibition with a combined oral contraceptive containing 1mg micronized 17 beta-estradio. Fertil Steril. 1993;60(4):616-9.
76. Grimes DA, Godwin AJ, et al. Ovulation and follicular development associated with three low-dose oral contraceptives: a randomized controlled trial. Ob-Gynecol. 1994;83(1):29-34.
77. Kim-Bjorklund T, Landgren BM, Johannisson E. Morphometric studies of the endometrium, the fallopian tube, and the corpus luteum during contraception with the 300 ug norethisterone (NET) minipill. Contraception 1991:43:459-74.
78. Parazzina F, Negri E, Ricci E, et al. Correlates of oral contraceptive use in Italian women. Contraception 1996,54:101-6.
79. Gianaroli L, et al. Preimplantation genetic diagnosis increases the implantation rate in human in vitro fertilization by avoiding the transfer of chromosomally abnormal embryos. Fertil Steril. Vol 68, No. 6, Dec l997.
CONTROVERSIES and CLARIFICATIONS
Appendix 3 (Refers to Page 3, Paragraph 2, of our original paper)
The scientific debate surrounding the question as to whether oral contraceptives (OCPs) are abortifacients has focused intently on scientific publications which might lend some insight into the phenomenon of breakthrough ovulation and the subsequent response of the endometrial lining. Of these, V. Chowdhury’s article has been called to the docket frequently to give its testimony.
We have reviewed his 1980 article “Escape ovulation in women due to the missing of low dose combination oral contraceptive pills” by V. Chowdhury et.al. (2) and have also been in personal correspondence with the authors. We have also reviewed a number of newer research articles on the subject of escape ovulation and ovarian activity on the combined oral contraceptive pills (see list of appendix references.) We would like to briefly discuss these below.
In brief, the 1980 Chowdhury article studied “ovulation” in 35 women who were previously sterilized and then asked to take a 30 ug ethinyl estradiol plus norethindrone acetate combination OCP. They were asked to “miss” 2 consecutive pills in a cycle, and then progesterone levels were measured at day 22 of the cycle, and endometrial biopsies were also obtained. Chowdhury found that 10 out of 35 women had progesterone levels greater then 4 ng/ml. He concluded that these 10 women had ovulated, based solely on its level of progesterone.
But, is a single serum progesterone level of greater than 4 ng/ml sufficient evidence to prove ovulation? Many authors have addressed this question. The answer is: “Clearly, No”. Let us look at one of these studies more closely; the 1982 article by Hull et.al. (6): “The value of a single serum progesterone measurement in the midluteal phase as a criterion of a potentially fertile cycle (“ovulation”) derived from treated and untreated conception cycles.” Hull looked specifically at attempts to determine ovulatory cycles by measuring serum progesterone in the midluteal phase (i.e. day 25+/-2 of cycle) of cycles that conceived. He studied conception cycles because those cycles which conceive are the only cycles where we can currently prove conclusively that an ovulation actually occurred; they are the only documentably proven ovulatory cycles. Let’s look at his findings:
“In an extended study a total of 21 untreated singleton conception cycles have been observed with a mean progesterone value of 40.7 nmol/ml (12.7 ng/ml), 95% confidence limits of 28-53 nmol/ml (8.8 to 16.7 ng/ml), and a range of 27 to 53 nmol/ml (8.5 to 16.7 ng/ml). This range was much narrower than for nonconception cycles (3 to 80 nmol/ml, 0.9 to 25.2 ng/ml), which extended significantly above as well as below the conception range, indicating that there is an optimal range for fertility with both an upper and a lower limit. The lower limit is of greater practical importance; and partly to allow for assay variation, we suggest it should be taken as 30 nmol/ml (9.4 ng/ml). It provided a clinically reliable criterion of potential fertility (“Ovulation”) in related studies. Our findings in treated conception cycles suggest that a higher value may be needed after treatment with clomiphene or gonadotropins because of the contribution from other stimulated follicles.” (6).
Hull defined the lower limit of progesterone produced in a cycle where ovulation was possible. Below that level of progesterone, ovulation does not happen. Other authors have suggested the same lower limit of 8-9 ng/ml of progesterone as the lowest limit of a potentially ovulatory cycle (1,3,10,15). Therefore, we corresponded with Dr. Chowdhury in order to obtain more precise information about the actual progesterone levels of his study participants. However, he replied that all the available information about that study was fully published in the paper, and he has no more detailed information than that which is already published. Therefore, we must conclude that we have no idea how many of his 10 patients actually were potentially ovulating (i.e. had a progesterone level high enough to support ovulation.) It is possible that if none of those 10 women had progesterone levels greater than 8 ng/ml, that none of them were actually ovulating. This renders the rest of his results essentially meaningless, because you cannot determine whether or not the endometrium was hostile to implantation in an ovulatory cycle on the OCP unless you determine that you actually HAVE an ovulatory cycle.
The second weakness of the Chowdhury article is the endometrial biopsy histology reporting. Chowdhury states: “The endometrial biopsy showed ‘hormone effect’ as reflected histologically by atrophic glands with excessively stimulated stroma”.
However, Mazur showed that excessive stromal hypertrophy was present in inadvertent endometrial biopsies performed in early gestation (8), and postulated that this was a necessary step in preparing the endometrium for implantation. Also, errors in histology can occur from sampling of the lower uterine segment instead of the fundus (12). Without more description of the actual histology obtained in Chowdhury’s biopsies, it is difficult to tell whether or not his specimens actually show “hostile endometrium”. (Of further interest is an article by Navot (11) who actually used supraphysiologic doses of Estrogen and progesterone to support the implantation and early pregnancy of women who were without any ovarian function of their own, but who had been recipients of IVF with donor embryos.)
Chowdhury further states: “In 5 out of 35 women in the first cycle treatment group and in 7 out of the 19 in the fourth cycle treatment group, the endometrium was so scanty that a suitable endometrial tissue sample could not be obtained.” However, there are other reasons as well why a tissue sample cannot be obtained, and it does not always mean “scanty endometrium.” In fact, frequently uterine fibroids, a retroflexed uterus, pain on the part of the patient, and operator inexperience are all reasons for insufficient tissue sampling. In fact, we are forced to conclude that in 14-35% of his data, the endometrial biopsy material is insufficient for meaningful interpretations.
Thus, the question of whether OCPs produce a “hostile endometrium” with breakthrough ovulations and in such instances are functionally chemical abortifacients remains an unanswered question for the following reasons:
1) Chowdhury’s study does not clearly identify a subgroup of patients on the OCP who are clearly ovulating on the OCP. A 4 ng/ml progesterone cutoff is inadequate to indicate ovulation, and his raw data is not available for further review at this time.
2) Even if available, a progesterone level > 9 ng/ml is only “permissive” of ovulation: i.e. a level < 9 ng/ml precludes ovulation, but a level > 9 mg/ml cannot distinguish reliably between ovulatory and nonovulatory cycles. This is because of significant contributions of progesterone production by luteinized unruptured follicles, which are follicles in the ovary which have not released the egg, yet still produce progesterone (6 and others. See literature on polycystic ovarian syndrome.).
3) Chowdhury’s endometrial biopsy are uninterpretable because of the lack of clear documentation of ovulation, and the large number of biopsies with no tissue obtained (i.e. 14-35% of his endometrial biopsies had no tissue.).
4) Improvements in ovulation detections were not utilized in the Chowdhury study (e.g. LH surge testing, ultrasound demonstration of ovulatory follicles or luteal phase endometrial thickening) limiting the study’s interpretation and utility.
However, the concept behind Dr. Chowdhury’s article is well worth repeating in the current era of availability of ultrasound assessment of ovarian function and evaluation of the endometrial lining, LH and FSH surge testing, and estradiol and progesterone assays. We would propose a new study to reexamine this issue, and are currently seeking support to implement this.
References, Appendix 3
1. Andoh, H. et al. “Endometrial dating in the conception cycle” Fertil.Steril. 58 (6) Dec. 1992 pp 1127-1130.
2. Chowdhury, V. et.al.; “Escape ovulation in women due to the missing of low dose combination oral contraceptive pills.” Contraception, Sept. 1980 vol. 22, no. 3.
3. Davis, O.K. et al. “Preparation of the endometrium for oocyte donation.” J. Asst. Repro. Genetics. 10(7) 1993. pp457-459.
4. Glissant, A. et al. “Ultrasound study of the endometrium during in vitro fertilization cycles”. Fertil.Steril. 44(6) Dec. 1985 pp. 786-790.
5. Hecht, B.R., et.al. “Luteal insufficiency: correlation between endometrial dating and integrated progesterone output in clomiphene citrate-induced cycles”. AJOG 163(6) Part 1, Dec., 1990. pp 1986-1991.
6. Hull, M.G.R., et.al. “The value of a single serum progesterone measurement in the midluteal phase as a criterion of a potentially fertile cycle (“ovulation”) derived from treated and untreated conception cycles”. Fertil. Steril. 37(3) Mar. 1982, pp. 355-360.
7. Killick S.R., et.al. “Ovarian activity in women taking an oral contraceptive containing 20 ug. ethinyl estradiol and 150 ug. desogestrel: effects of low estrogen doses during the hormone-free interval”. AJOG 179(1) Jul. 1998, pp. s18-s24.
8. Mazur, M.T. et.al. “Endometrial biopsy in the cycle of conception: histologic and lection histochemical evaluation”. Fertil. Steril. 51(5) May 1989, pp, 764-769.
9. Michalas, S. et.al. “A flexible protocol for the induction of recipient endometrial cycles in an oocyte donation programme”. Hum. Repro. 11(5), pp. 1063-1066, 1996.
10. Navot, D., et.al. “Hormonal manipulation of endometrial maturation”. JC&E68(4), pp. 801-807, 1989.
11. Navot, D., et.al. “Artificially induced endometrial cycles and establishment of pregnancies in the absence of ovaries”. NEJM314(13), 1986, pp. 806-811.
12. Psychoyos, A. “Uterine receptivity to nidation”. Annals of NY Academy of Science, 476, pp. 36-42, 1986.
13. Rosenfeld, D.L. “A comparison of endometrial histology with simultaneous plasma progesterone determinations in infertile women”. Fertil.Steril. 27(11) Nov. 1976, pp. 1256-1266.
14. Rossmanigh, W.G., et.al. “A comparative randomized trial on the impact of two low-dose oral contraceptives on ovarian activity, cervical permeability and endometrial receptivity”. Contrac. 1997; 56:23-30.
15. Shoupe, D., et.al. “Correlation of endometrial maturation with four methods of estimating day of ovulation”. Obstet. Gynec. 73(1), Jan. 1989, pp. 88-92.
Copies of this paper may be obtained upon request from ProLife Obstetrician, PO Box 81, Fennville, Michigan 49408, or by e-mail from firstname.lastname@example.org