Risks and Benefits of Oral Contraceptives: Will Breast Cancer Tip the Balance?
DAVID C. G. SKEGG
Concern about breast cancer is the main cloud hanging over oral contraceptives. When they were first marketed three decades ago, some hoped they might reduce the risk of breast cancer. The Puerto Rico contraceptive study, instigated by Gregory Pincus in 1961, was intended to explore the “anticancer effect on the breast and genital system” of using the pill (Potts et al., 1982). The recognition that oral contraceptives reduced the risk of benign breast disease fueled the hope that they would also protect against breast cancer. The great majority of studies, however, have shown no alteration of risk; indeed, several recent investigations have suggested that oral contraceptives may increase the risk of breast cancer in certain groups of women (Schlesselman, 1989; Thomas, 1989; Mann, 1990). The current decline in oral contraceptive use by American women has been attributed to media publicity about these positive studies (Contraception Report, 1990).
Because breast cancer is so common, with a cumulative incidence rate (before age 75) of about 9 percent in the United States (Muir et al., 1987), any general increase in risk associated with a widely used method of contraception would be of serious concern. Breast cancer also happens to be a disease that women and their families particu-
David C. G. Skegg, University of Otago Medical School, Dunedin, New Zealand.
larly fear, so any increase in risk might carry a disproportionate weight when choices about contraception are being made. Nevertheless, women and their physicians need to weigh up all of the benefits and risks of oral contraceptives, and to decide whether any increased risk of breast cancer is likely to tip the balance. To do this, it is necessary to consider the range of breast cancer effects that might plausibly emerge over the next few decades.
ESTABLISHED BENEFITS AND RISKS
The outstanding benefit of oral contraception is its prevention of unplanned pregnancy—with such a high degree of effectiveness, convenience, and reversibility (Ory et al., 1983). Future historians will attempt to assess the pill's enormous range of influences on twentieth-century society. From a medical point of view, the effectiveness of the pill is of greatest significance to women in developing countries. Whereas maternal mortality is something many of us read about in Victorian novels, it is still a grim reality in developing countries. Each year pregnancy and childbirth claim the lives of at least half a million women (Royston and Armstrong, 1989). A new World Health Organization report estimates that, in Bali, a woman has a 1-in-32 lifetime risk of dying from pregnancy-related causes; in Bangladesh the risk is 1-in-26, and in parts of rural Africa (where high maternal mortality rates coexist with high fertility) it may be as high as 1 in 15 (Royston and Armstrong, 1989). By preventing unwanted pregnancies and enabling women to space births, the pill contributes not only to maternal health but also to the survival and health of children.
Apart from the prevention of pregnancy, oral contraceptives have a wide range of benefits and risks. The noncontraceptive benefits of combined oral contraceptives include reductions in the incidence of menstrual problems (such as dysmenorrhea and menorrhagia), iron-deficiency anemia, pelvic inflammatory disease, functional ovarian cysts, benign breast disease, epithelial ovarian cancer, and endometrial cancer (Prentice and Thomas, 1987; Population Information Program, 1988; Vessey, 1990). The pill's strong protective effects against cancers of the ovary and endometrium are especially important because the protection appears to persist in ex-users for at least 15 years after stopping oral contraception (Cancer and Steroid Hormone Study, 1987a,b).
The most important adverse effects of combined oral contraceptives are their propensity to cause cardiovascular events such as myocardial infarction, thrombotic stroke, hemorrhagic stroke, and venous thrombosis and embolism (Stadel, 1981; Prentice and Thomas, 1987).
Fortunately, these complications are rare and largely confined to current users of oral contraceptives. The increased risk of myocardial infarction is found mainly in women with other risk factors, such as cigarette smoking. The demonstration of these other factors in epidemiological studies has led to much more careful screening of candidates for oral contraception, with the result that many of these catastrophic occurrences are now avoided. It should also be noted that our estimates of cardiovascular risks are based on studies that were carried out in the 1970s, and it is suspected (although not yet proven) that modern low-dose oral contraceptives carry lower risks.
Long-term use of the pill increases the risk of hepatocellular adenoma (Rooks et al., 1979) and (probably) hepatocellular carcinoma (Forman et al., 1986; Neuberger et al., 1986), but these conditions are rare in women of childbearing age in the United States. More worrying is an association between long-term use of oral contraceptives and invasive cancer of the cervix (Prentice and Thomas, 1987; Hulka, 1989; Schlesselman, 1989). It is still unclear whether this association reflects a causal relationship or incomplete adjustment for the confounding influences of the sexual behavior of women and their male partners (Swan and Petitti, 1982).
There are many other suspected beneficial and adverse effects of oral contraception that require further research (Prentice and Thomas, 1987; Population Information Program, 1988; Vessey, 1990). The suggestion that use of the pill might increase the risk of contracting human immunodeficiency virus (HIV) infection, based on a study of prostitutes in Nairobi (Simonsen et al., 1990), has not been confirmed.
BALANCE OF BENEFITS AND RISKS
Any comparison of the risks and benefits of oral contraceptives needs to be specific to particular countries or groups of countries, because the balance sheet will be affected by both the underlying incidence of diseases and the risk of maternal mortality. In those developing nations where maternal mortality is high, the effectiveness of the pill in preventing pregnancy will be of overwhelming importance. Even when attention is focused on a single country, the balance of benefits and risks will be different for different groups of women. Thus, the balance will be different for younger and older women, for smokers and nonsmokers, and so on.
Despite such complexities, several authors have tried to draw up balance sheets of benefits and risks (Tietze et al., 1976; Ory et al., 1983; Prentice and Thomas, 1987; Vessey, 1990). Their approaches have inevitable limitations. First, no value is placed on avoidance of
the grief of unwanted pregnancy. Pregnancy is counted only as a possible cause of morbidity and death. Second, attention is generally focused on mortality: how could one compare the alleviation of dysmenorrhea in 1,000 women with the causation of a stroke in one woman? On the other hand, counting mortality enables direct comparisons to be made, and the approach can perhaps be defended on the grounds that “death is unequivocal and of overriding importance to the individual” (Doll, 1973).
Vessey recently described a simplified model of mortality that compares users of oral contraceptives and of condoms in the United Kingdom (Vessey, 1990). He assumes that a cohort of one million women start to use combined oral contraceptives at age 16 and that another cohort of one million women decide to rely on the condom from the same age. Both groups of women continue with their chosen method of birth control until the age of 35, when they (or their partners) are sterilized. Follow-up continues to the age of 50. If it is assumed that oral contraceptives do not affect the risk of breast or cervical cancer, the women in the pill cohort do appreciably better, with 1,134 fewer deaths overall (1,240 fewer if lower cardiovascular risks are assumed). The most striking effect is the prevention of 1,400 deaths from ovarian cancer in the oral contraceptive cohort.
If it is assumed that the association between the pill and cervical cancer is causal and that it persists in ex-users, Vessey estimates that the mortality advantage in the oral contraceptive cohort will be reduced by 764 additional deaths from cervical cancer. He then incorporates the positive findings of the recent U.K. National Case-Control Study of oral contraceptives and breast cancer (U.K. National Case-Control Study Group, 1989). If users of the pill have an increased risk of breast cancer up to the age of 35 but not at older ages, he estimates that their mortality advantage will be reduced by 311 additional deaths from breast cancer. But if it is assumed, as some have suggested, that the increased risk of breast cancer observed in the U.K. National Study will persist beyond age 35, the picture looks very different. Assuming that a relative risk of 1.75 in long-term users persists until age 50, the additional number of deaths from breast cancer in the oral contraceptive cohort will not be 311 but 4,157. Hence, a persisting risk of breast cancer would certainly tip the balance against oral contraceptives.
ORAL CONTRACEPTIVES AND BREAST CANCER
How likely or unlikely is this pessimistic scenario? It must be regarded as very unlikely on the evidence to date because, as noted
in other reviews (Schlesselman et al., 1989; Thomas, 1989), the results of case-control and cohort studies in middle-aged women have been reassuringly negative. Some commentators, however, have suggested that a new phenomenon may be emerging (McPherson et al., 1986; Lund, 1989).
Interpreting the positive findings in several recent case-control studies is difficult for two main reasons. The first problem is one of shifting goalposts: the positive subgroups keep changing. For example, Pike's group in Los Angeles reported in 1981 that women who used the pill before their first full-term pregnancy had an increased risk of breast cancer at a young age (Pike et al., 1981). Oral contraceptive use after the first pregnancy was not associated with any change in risk. In 1983, the same group reported on an expanded analysis (Pike et al., 1983). There was a strong association with use of the pill before age 25: the relative risk was estimated to be 2.0 for four to six years of use and 4.9 for more than six years of use before age 25. The previous association with use before the first pregnancy was now attributed to a positive correlation between this variable and use before age 25. Attention was subsequently focused again on use before the first pregnancy by the results of a case-control study conducted in Britain by McPherson and colleagues (McPherson et al., 1983, 1987). They found a strong association with use before the first pregnancy but no effect of use after it—indeed, there was a suggestion that use occurring only after the first pregnancy might be protective.
Pike, McPherson, and Vessey were then instrumental in setting up the U.K. National Case-Control Study, with 755 cases under age 36 and an equal number of controls (U.K. National Case-Control Study Group, 1989). The results published in 1989 showed a highly significant trend in risk of breast cancer with total duration of pill use. But not only were the relative risk estimates much closer to 1.0 than in the previous studies; there was now no greater effect of use before the first pregnancy than after the first pregnancy. The results were presented clearly, but many readers apparently did not appreciate that the authors' previous hypothesis (about a risk confined to use at a specific time in early reproductive life) had been rejected.
Clearly, use before the first pregnancy, use before age 25, and total use will all be correlated, and many studies will have insufficient power to distinguish their effects with confidence. There is a danger, however, in focusing attention on the subgroup in each study that happens to give the highest relative risk estimate. This source of bias is illustrated by the summary table in a recent British review (Chilvers and Deacon, 1990), which shows only the subgroup in each study that gave the most positive results (ignoring the fact that negative results were obtained in some subgroups listed for other studies).
The second problem in interpreting recent findings is that in several of the best investigations, such as the U.K. National Study (U.K. National Case-Control Study Group, 1989), the relative risk estimates are close to 1.0. With relative risks of this order, epidemiologists find it difficult to exclude the possible influences of bias, confounding, and chance (Skegg, 1988). With regard to precautions taken to minimize and assess potential sources of bias, the U.K. National Study is the most adequate case-control study so far conducted; nevertheless, some problems remain. For example, selection (or nonresponse) bias is a possibility because only 72 percent of eligible cases could be interviewed (16 percent had died). Examination of general practice case-notes showed that the women with breast cancer who were not interviewed had, on average, significantly less oral contraceptive use than those who were interviewed. Data presented in the paper suggest that the relative risks should probably be scaled down by about 20 percent to allow for this bias, which would bring them even closer to 1.0.
The possibility of modification of the relative risk by confounding factors is always present in studies of this issue. Adjustments are made for known confounding variables but, because the main causes of breast cancer are unknown, unsuspected factors could be associated with both choosing oral contraception and the risk of developing breast cancer.
Cohort studies are free of some, although not all, of the sources of bias that can affect case-control studies, and definitive conclusions may be possible only with the completion of cohort studies containing large numbers of young women—such as the new Nurses' Health Study. The best synthesis of the evidence so far available from case-control and cohort studies is that oral contraceptives do produce a modest increase in the risk of developing breast cancer at a young age (up to about 35 years). This was the provisional conclusion reached by Sir Richard Doll, in summing up a recent conference at which data from most of the major studies were presented (Doll, 1990). Doll concluded that use of oral contraceptives produces no material increase in the risk of developing breast cancer after the age of about 45. With regard to the 35- to 44-year-old age group, he thought that it is “reasonable to postulate that there is some tailing off of the effect” occurring at younger ages.
AN EFFECT IN YOUNG WOMEN
If any adverse effect of oral contraceptives is confined to the risk of breast cancer developing at a young age, the situation is close to
the more optimistic scenario considered by Vessey in his risk-benefit assessment (Vessey, 1990). Can it be confidently assumed, however, that the effect observed in young women will not extend in the future to the older ages at which breast cancer is common? It is essential to consider the possible explanations for the different results obtained recently in young and older women. Four possible explanations have differing implications for the future.
First, there could be a specific effect on the risk of breast cancer at a young age, which either tails off or reverses at older ages. This is the hypothesis advanced by Doll (1990), and it is plausible because there is known to be a “crossover” in the effects of some other risk factors for breast cancer with age (Janerich and Hoff, 1982; Ron et al., 1984).
Second, there could be an effect of recent use of the pill, which either tails off or reverses after some years. Here the best analogy would be with the initial influence of pregnancy. A full-term pregnancy appears to be followed by a short-term increase in the risk of breast cancer, which, after some years, is superseded by the protective effect of parity (Bruzzi et al., 1988). La Vecchia proposed that oral contraceptives might have an effect similar to that of pregnancy, although the evidence supporting this hypothesis is limited (La Vecchia, 1990).
Third, there could be an effect of use early in reproductive life, which has become common only among the most recent cohorts of women. This hypothesis led McPherson and others to suggest the possibility of a major epidemic of breast cancer as women in these cohorts age (McPherson et al., 1986). The results of the recent U.K. National Study provide no support for this explanation, because women who started using the pill at a young age (e.g., in their teens) were not found to be at higher risk than those who started later (U.K. National Case-Control Study Group, 1989). The final results from a large case-control study in New Zealand also provide evidence against an adverse effect of starting oral contraceptives before 20 years of age (Paul et al., 1990).
Fourth, there could be an effect of the particular formulations of the pill used by recent cohorts of women. Such an effect might also be expected to persist as these women move into older age groups. Again, the U.K. National Study provides no support for this explanation because relative risks were found to be higher for the older (high-estrogen) combined pills (U.K. National Case-Control Study Group, 1989). Nevertheless, it remains possible that some of the discrepancies among studies reflect geographic and temporal patterns of use of pills with particular steroid combinations or dosages.
Although doubt may persist until the present generation of young women moves into middle age and beyond, the first of these four explanations seems the most plausible at present.
All contraceptives cause breast cancer to the extent that delaying a woman's first pregnancy (and probably reducing her total number of pregnancies) increases her risk of breast cancer (Layde et al., 1989). The cloud hanging over oral contraceptives is the suggestion that these contraceptives may cause a specific increase in risk, at least in young women. The relationship between the pill and breast cancer appears complex and will not be clarified by a focus only on positive subgroups. If future studies are not to produce results that may be misleading from a public health viewpoint, they should cover the full range of ages of women who have used oral contraceptives (including women who have passed the menopause).
More research is needed before we can be certain that a risk of breast cancer does not outweigh the noncontraceptive benefits of oral contraception (especially in preventing ovarian cancer). Present evidence offers grounds for cautious optimism: breast cancer is not likely to tip the balance against oral contraceptives. The cloud over the pill may also have a silver lining if a better understanding of its relationship with breast cancer can lead to the design of hormonal contraceptives that reduce the risk of breast cancer, as well as cancers of the ovary and endometrium. The primary prevention of breast cancer will then be within our grasp.
Bruzzi, P., E. Negri, C. La Vecchia, et al. 1988. Short-term increase in breast cancer risk following a full-term pregnancy British Medical Journal 297: 1096-1098
Cancer and Steroid Hormone Study. 1987a. Combination oral contraceptive use and the risk of endometrial cancer Journal of the American Medical Association 257: 796-800
Cancer and Steroid Hormone Study. 1987b. The reduction in risk of ovarian cancer associated with oral-contraceptive use. New England Journal of Medicine 316: 650-655
Chilvers, C. E. D., and J. M. Deacon. 1990. Oral contraceptives and breast cancer. British Journal of Cancer 61: 1-4
Contraception Report. 1990. Pp. 4-5 in The Media and Contraception, Vol. 1 Morris Plains, N.J.:
Doll, R. 1973. Monitoring the National Health Service. Proceedings of the Royal Society of Medicine 66: 729-740
Doll, R. 1990. What conclusions can we reach and how do we best assist women to
make informed choices regarding their use of oral contraceptives? Pp. 395-400 in Oral Contraceptives and Breast Cancer, R. D. Mann, ed. Park Ridge, N.J.: Parthenon Publishing.
Forman, D., T. J. Vincent, and R. Doll. 1986. Cancer of the liver and the use of oral contraceptives. British Medical Journal 292: 1357-1361
Hulka, B. S. 1989. Hormonal contraceptives and risk of cervical cancer. Pp. 84-96 in Safety Requirements for Contraceptive Steroids, F. Michal, ed. Cambridge: Cambridge University Press.
Janerich, D. T., and M. B. Hoff. 1982. Evidence for a cross-over in breast cancer risk factors. American Journal of Epidemiology 116: 737-742
La Vecchia, C. 1990. Oral contraceptives and breast cancer: Update of an Italian case control study. Pp. 245-251 in Oral Contraceptives and Breast Cancer, R. D. Mann, ed. Park Ridge, N.J.: Parthenon Publishing.
Layde, P. M., L. A. Webster, A. L. Baughman, et al. 1989. The independent associations of parity, age at first full term pregnancy, and duration of breastfeeding with the risk of breast cancer Journal of Clinical Epidemiology 42: 963-973
Lund, E. 1989. Re: “Potential for bias in case-control studies of oral contraceptives and breast cancer” (letter). American Journal of Epidemiology 130: 1066-1068
Mann, R. D., ed. 1990. Oral Contraceptives and Breast Cancer. Park Ridge, N.J.:
McPherson, K., A. Neil, M. P. Vessey, and R. Doll. 1983. Oral contraceptives and breast cancer (letter). Lancet 2: 1414-1415
McPherson, K., P. A. Coope, and M. P. Vessey. 1986. Early oral contraceptive use and breast cancer: Theoretical effects of latency. Journal of Epidemiology and Community Health 40: 289-294
McPherson, K., M. P. Vessey, A. Neil, R. Doll, L. Jones, and M. Roberts. 1987. Early oral contraceptive use and breast cancer: Results of another case-control study. British Journal of Cancer 56: 653-660
Muir, C., J. Waterhouse, T. Mack, J. Powell, and S. Whelan, eds. 1987. Cancer Incidence in Five Continents, vol 5. Lyon: International Agency for Research on Cancer.
Neuberger, J., D. Forman, R. Doll, and R. Williams. 1986. Oral contraceptives and hepatocellular carcinoma. British Medical Journal 292: 1355-1357
Ory, H. W., J. D. Forrest, and R. Lincoln. 1983. Making choices: Evaluating the health risks and benefits of birth control methods. New York:
Paul, C., D. C. G. Skegg, and G. F. S. Spears. 1990. Oral contraceptives and risk of breast cancer. International Journal of Cancer 46: 366-373
Pike, M. C., B. E. Henderson, J. T. Casagrande, I. Rosario, and G. E. Gray. 1981. Oral contraceptive use and early abortion as risk factors for breast cancer in young women. British Journal of Cancer 43: 72-76
Pike, M. C., B. E. Henderson, M. D. Krailo, A. Duke, and S. Roy. 1983. Breast cancer in young women and use of oral contraceptives: Possible modifying effect of formulation and age at use. Lancet 2: 926-930
Population Information Program. 1988. Lower-dose pills. Population Reports, SeriesA, No. 7. Baltimore, Md.
Potts, M., P. J. Feldblum, I. Chi, W. Liao, and A. Fuertes-de La Haba. 1982. The Puerto Rico oral contraceptive study: An evaluation of the methodology and results of a feasibility study British Journal of Family Planning 7: 99-103
Prentice, R. L., and D. B. Thomas. 1987. On the epidemiology of oral contraceptives and disease. Advances in Cancer Research 49: 285-401
Ron, E., F. Lubin, and Y. Wax. 1984. Re: “Evidence for a cross-over in breast cancer risk factors” (letter). American Journal of Epidemiology 119: 139-141
Rooks, J. B., H. W. Ory, K. G. Ishak, et al. 1979. Epidemiology of hepatocellular adenoma. The role of oral contraceptive use. Journal of the American Medical Association 242: 644-648
Royston, E., and S. Armstrong, eds. 1989. Preventing Maternal Deaths. Geneva: World Health Organization.
Schlesselman, J. J. 1989. Cancer of the breast and reproductive tract in relation to use of oral contraceptives. Contraception 40: 1-38
Simonsen, J. N., F. A. Plummer, and E. N. Ngugi. 1990. HIV infection among lower socioeconomic strata prostitutes in Nairobi AIDS 4: 139-144
Skegg, D. C. G. 1988. Potential for bias in case-control studies of oral contraceptives and breast cancer. American Journal of Epidemiology 127: 205-212
Stadel, B. V. 1981. Oral contraceptives and cardiovascular disease. New England Journal of Medicine 305: 612-618, 672-677
Swan, S. H., and D. B. Petitti. 1982. A review of the problems of bias and confounding in epidemiologic studies of cervical neoplasia and oral contraceptive use. American Journal of Epidemiology 115: 10-18
Thomas, D. B. 1989. The breast. Pp. 38-68 in Safety Requirements for Contraceptive Steroids, F. Michal, ed. Cambridge: Cambridge University Press.
Tietze, C., J. Bongaarts, and B. Schearer. 1976. Mortality associated with the control of fertility. Family Planning Perspectives 8: 6-14
U.K. National Case-Control Study Group. 1989. Oral contraceptive use and breast cancer risk in young women. Lancet 1: 973-982
Vessey, M. P. 1990. An overview of the benefits and risks of combined oral contraceptives Pp. 121-132 in Oral Contraceptives and Breast Cancer, R. D. Mann, ed. Park Ridge, N.J.: Parthenon Publishing.