Oral contraceptives, which have been in use now for 30 years in the United States, are a highly effective and popular means of contraception. Data from 1988 indicate that at least 10.7 million American women currently use this method of contraception; 80 percent of 35year-olds use, or have used, oral contraceptives. There are 17 different oral contraceptive formulations now available in the United States, which are marketed under 29 brand names.
Behind these impressive statistics are noteworthy trends: both the patterns of oral contraceptive use and the chemical formulation of the pills have been changing. The trends have been toward smaller and smaller doses of active hormones and increasing use by younger and younger women to prevent a first pregnancy, rather than to space successive pregnancies. Concomitant with the trend toward use at younger ages is the likelihood that today's young women will use oral contraceptives longer than previous, older users.
The introduction and dissemination of oral contraceptives in the United States have been superimposed on steadily rising breast cancer incidence rates. Breast cancer accounts for more new cases of cancer among American women than cancer of any other organ, and it has become more common through the twentieth century. One in nine women develops the disease sometime in her life, and 1 in 18 can expect to die from it.
Both pre- and postmenopausal endogenous hormones almost certainly play an important role in breast cancer, and their levels prob-
ably moderate the greater or lesser risk of women of different ages and with different reproductive histories. Therefore, it is biologically plausible that exogenous hormones, such as the synthetic estrogens and progestins found in oral contraceptives, have analogous effects. One should be prepared to find qualitatively and/or quantitatively different effects of exogenous hormones in older and younger women, and among women at any age with different reproductive histories.
Human studies of the effects of oral contraceptive use on breast cancer incidence have been complicated by the relatively short history of such use and a lack of stability of oral contraceptive formulations and usage patterns. Studies have collected and analyzed information only about relatively short-term exposures in women who started use as teenagers and about longer-term exposures among those who began to use the pill when they were in their 20s or early 30s. Both groups of women have had varied patterns of starting and stopping oral contraceptive use, and a majority have used two or more different formulations. (Oral contraceptives can vary in the types, amount, and ratios of hormones they contain.)
The utility of results of existing epidemiological studies in defining the relationship of oral contraceptives and breast cancer depends on whether they include exposures and follow-up intervals that are biologically feasible for identifying associations of both initiating and promoting events. Ideally, they should provide information about the dose, consistency, and duration of use, as well as the timing relative to reproductive physiological events (i.e., menarche, pregnancy, menopause) that perturb endogenous hormone levels.
There is considerable agreement about the consistency of results of the existing, prospective epidemiological studies. Thus far, the women in those studies have had no increases in breast cancer as they entered early postmenopausal years. Relatively few (about 10) additional years of follow-up are needed to determine whether the findings hold for later postmenopausal years. This follow-up, as well as a dedicated effort to prepare the results of existing cohort studies for pooled analysis, should be enough to confirm the existing consensus and put statistical limits on the upper bounds of measured safety with respect to the risk of breast cancer. There may also be sufficient data from this effort to determine whether oral contraceptive exposure of these virtually first users has had any impact in their postmenopausal years. The lack of an association between oral contraceptive use and postmenopausal breast cancer in these women would be an encouraging finding that will need to be confirmed
among women who have long-term exposure from an early age. Studies of the impact of oral contraceptive use on postmenopausal breast cancer rates are imperative, because the majority of all breast cancer cases occur in the postmenopausal years.
In contrast to studies of the first women to use oral contraceptives, studies of the impact of more recent patterns of use (i.e., long-term exposures beginning early in life and before a full-term pregnancy) are in progress. Given that current low-dose oral contraceptive formulations will constitute the major proportion of all future exposures, and that studies to date have largely involved older, higher-dose formulations, studies of current (i.e., late 1980s and early 1990s) cohorts will be critical. Information about timing and duration of exposures will be particularly important and is likely to be equally—if not more —important than dosage in efforts to unravel early and late carcinogenic effects.
The committee considered efficient ways to obtain information about exposure, timing, and duration, and still ensure that any harmful effects will be detected in as short a time as possible. How long it will take to get the information is being dictated as much by the basic biology of breast cancer as by research design. The effort to untangle the relationship, if any, between oral contraceptives and breast cancer will be long and slow; to be successful, it must be comprehensive. Comprehensiveness requires that it be tightly linked to progress made in the overall understanding of the biology of the breast and the pathogenesis of the disease.
Data obtained from studies of different experimental mammals, although conflicting, suggest that some of the steroids used in oral contraceptives are capable of inducing tumors. A formal, four-step cancer risk assessment methodology (i.e., hazard identification, hazard characterization, exposure assessment, and risk characterization) has been used for more than 30 years by regulatory agencies. This risk assessment strategy extrapolates observed risk in animals to humans. The availability of human data from epidemiological studies allows for a desirable modification of the traditional approach.
The effect, if any, of oral contraceptives on breast cancer is perhaps the last substantial gap in knowledge of the side effects of the pill. The committee thus offers recommendations to address four policy issues related to diminishing that gap: (1) maintaining surveillance, (2) developing a broader array of contraceptives, (3) assessing knowledge for use in clinical practice, and (4) filling gaps in biological and epidemiological knowledge. The committee's recommendations are summarized in Table 1.
TABLE 1 Recommendations Made by the IOM Committee on the Relationship Between Oral Contraceptives and Breast Cancer
Developing a broader array of contraceptives
Assessing knowledge for use in clinical practice
Filling gaps in biological and epidemiological knowledge
With oral contraceptives in widespread use, epidemiologists must maintain appropriate surveillance through the next 20 to 40 years to follow the effects of current and new formulations of the pill when used from an early age. They must also establish studies of risks and benefits to ensure the relative safety of new formulations. These future studies must (1) involve many thousands of cases and controls to permit statistically valid analyses of subgroups of users (e.g., individuals who used the pill prior to their first pregnancy), and (2) follow sufficient numbers of women for enough years to measure how long any possible effect persists after discontinuing oral contraceptive use. These are studies of daunting size and duration; nevertheless, they must be undertaken—and the necessary resources must be set aside to conduct them—to protect the health of American women.
The committee calls attention to the opportunity for international, cooperative research, owing to the complementarity of European and North American research data. There is enough variation in both exposure to oral contraceptives and opportunities for surveillance among the countries gathering data that a coordinated future research plan offers worldwide benefits. It is also important to note that some oral contraceptive formulations that are likely to become available in the United States over the next five years are already in use in a number of countries worldwide, including many European nations.
Further, the committee recommends that the Food and Drug Administration consider premarketing and postmarketing requirements as an integrated whole. There is a need to devise ways of spreading the investment already required by manufacturers, and ultimately paid for by users, in such a way as to maximize the information on safety that will result.
DEVELOPING A BROADER ARRAY OF CONTRACEPTIVES
The success and popularity of oral contraceptives as a means of avoiding conception are both a tribute to their efficacy and a reflection of the paucity of effective alternative means, particularly for young women. The committee underscores the urgent need for research and development of a broader array of contraceptives, including more effective barrier methods and nonsteroidal methods. It is not feasible to wait for resolution of the uncertainty about oral contraceptives as a potential risk factor for breast cancer before energiz-
ing research and development on alternative methods of contraception. In the face of uncertainty, those couples wishing to eschew oral contraceptives in favor of other methods of contraception must be served—and sooner, not later. Unless immediate steps are taken to develop a broader array of safe and effective contraceptives, the choice of contraceptives in the United States in the next century will not differ appreciably from the choice available today.
ASSESSING KNOWLEDGE FOR APPLICATION IN CLINICAL PRACTICE
The committee finds that, at the outset of the 1990s, the best available knowledge about oral contraceptives and breast cancer does not support any fundamental change in clinical practice with respect to use of the pill. This finding is subject to an important caveat: the knowledge base must be reassessed at regular intervals, in light of new research results (e.g., those forthcoming from the World Health Organization and the National Cancer Institute). The committee recommends that, three years hence, and at regular intervals thereafter, an NIH consensus conference reassess recommendations for oral contraceptive use based on reevaluation of the relationship between oral contraceptives and breast cancer. Two institutes in particular, the National Institute of Child Health and Human Development and the National Cancer Institute, would be appropriate lead institutes to judge when consensus conferences should be held. Current understanding of health benefits, as well as health risks, of oral contraceptives are described in detail in Appendix E and summarized in Chapter 4.
New research results should be periodically assessed for evidence of short- and long-term safety and carefully analyzed to provide clinicians with current information on the safest oral contraceptive use for younger and older women. There is also a need for continuous improvement of the dissemination of new and existing information about oral contraceptives and breast cancer among health care providers and women who use oral contraceptives. Multiple channels, through which providers and women can receive up-to-date information, must be opened and maintained. The committee emphasizes the obligation of all health care providers to offer adequate information to women seeking contraception and to counsel them regarding the current state of ambiguity with respect to the relationship between oral contraceptives and breast cancer. Only then can fully informed choices be made in clinical practice.
FILLING GAPS IN BIOLOGICAL AND EPIDEMIOLOGICAL KNOWLEDGE
The committee believes that a multidisciplinary research approach is necessary to resolve the relationship between oral contraceptives and breast cancer. Research on a broad front could well produce insights that would lead to some reduction in the incidence of breast cancer, in the same way that research throughout the past three decades has led to insights and changes of lifestyle that are today believed to be reducing the incidence of heart disease in America. With no single, “magic bullet” likely to prevent or cure breast cancer, the committee recommends a broad program of expanded studies on breast function and pathophysiology from the perspectives of endocrinology, biochemistry, molecular genetics, nutrition, cytology, and tissue culture.
Basic research in the biology of the breast is a priority. Research is especially needed on the transition from normal to abnormal growth, including, for example, research on the role of local growth factors and newly detected extracellular matrix proteins, estrogen metabolism, and oncogenes. As basic resources for this research, in vitro model systems that employ normal breast tissue are needed, in which the effects of oral contraceptive steroids can be directly examined.
Epidemiological research priorities include a large, case-control (retrospective) study with sufficient statistical power to resolve the assessment of small increases in relative risk of breast cancer. Such a study by the National Cancer Institute is in progress, involving primarily women below the age of 45. A study of postmenopausal women —who experience the bulk of breast cancer—is needed to elucidate the effects of (1) oral contraceptive use, and (2) oral contraceptive use followed by hormone replacement therapy. As oral contraceptive formulations and use patterns of the pill change, future case-control studies will be required. The launching of new cohort (prospective) studies is equal in priority to the conduct of case-control studies. There are no substitutes for these classical epidemiological approaches to resolving the relationship of oral contraceptives and breast cancer.
Biological markers, as they become more generally available, should be incorporated into epidemiological protocols. Consistent with a multidisciplinary approach to better understanding of the causes of breast cancer, there is a need for studies using biological markers within the context of epidemiological study designs, using innovative as well as traditional research tactics.
The expensive, complex program of studies necessary to reveal the relationship (if any) between oral contraceptives and breast cancer will inevitably draw on funding from several sources. The cost of such action must be weighed against the sobering costs of inaction: increased suffering and loss of life; the cost of treating breast cancer; and the cost of mammography screening programs, which lead to earlier detection but are not primary prevention. With oral contraceptives now being used by an estimated 10.7 million American women, any effect on breast cancer would be highly significant in terms of public health. Furthermore, any proven relationship or, what is equally important, any perceived relationship when one does not exist, would also have a marked effect on contraceptive practice in the United States.