Breast cancer and the pill

In the UK 11% of women will develop breast cancer by age 85. Breast cancer is the most frequently diagnosed cancer in women.¹ However, compared with lifetime risk, the age-specific risk of breast cancer is low. At age 50 the risk is 2% and at 25 it is 0.05% (1 in 2000). Whatever the actual risk, the notion of getting breast cancer is terrifying, and women overestimate the likelihood of acquiring and dying from this disease by a factor of 25.² Less than 5% of deaths in women are attributed to breast cancer; more than half of all women die of cardiovascular disease. Public anxiety about breast cancer tends to be fuelled by the media. Qualitative research suggests that cancer stories are overreported in the popular press, and both broadsheets and tabloid newspapers are 70% more likely to publish bad news than good news. They are also more inclined to report observational studies than randomized trials.³ In American magazines throughout the 1990s there was disproportionate reporting of breast cancer in a 2:1 ratio to cardiovascular disease.⁴

RISK FACTORS FOR BREAST CANCER

The aetiology of breast cancer is multifactorial, with genetic, environmental and reproductive factors interacting in a complex way. The impact of environmental and reproductive factors depends on the age of the woman.¹ Breast cancers can be classified by age at presentation as (a) reproductive-age cancers, occurring in women under the age of 40; (b) perimenopausal cancers, in women between 40 and 55, and (c) postmenopausal cancers, which account for the majority of breast cancers.¹ Postmenopausal cancers are oestrogen receptor positive and unrelated to smoking; reproductive-age cancers are less likely to be oestrogen receptor positive and smoking is cited as a risk.

Reproductive risk factors for breast cancer include nulliparity, delayed first full-term pregnancy, early menarche and late menopause.⁵ These also seem to be risk factors for ovarian and endometrial cancer. The common thread linking all these factors is lengthy exposure of sensitive tissue to cyclical hormonal stimulation driven by incessant ovulation. An intriguing observation is that the combined pill, an ovulation suppressant, reduces the risk of both endometrial and ovarian cancer but not breast cancer. This lack of protective effect may be because environmental factors are more important in the causation/promotion of breast cancer than in ovarian cancer. Indeed the rising breast cancer risk since the 1940s in developed countries is attributed to lifestyle factors.

Environmental and lifestyle factors include smoking. It is highly likely that breast cancer in young women is affected by smoking patterns. A study published last year⁶ suggests that a nulliparous woman smoking 20 cigarettes a day for 20 years increases her risk of breast cancer seven-fold. Young women who smoke in the 5 years from menarche likewise show an increased risk, and this is probably a direct effect on breast tumour oncogenesis rather than a hormone-mediated effect. Other lifestyle effects may be mediated via ovarian hormones; for example, high physical activity and low caloric intake reduce the concentrations of ovarian steroids and may thereby lessen the risk of breast cancer.

Ovarian steroids

There is biological evidence for the idea that oestrogens are a driver in breast cancer pathogenesis, and oestradiol is the oestrogen most likely to be implicated. Breast cancer is a hundred times more common in women than in men, occurs after puberty and many breast tumours have oestrogen receptors.⁷ The role of progesterone is uncertain. Mitotic activity in breast tissue does increase in the luteal phase, and progesterone may be responsible for this.⁸

Pregnancy, parity and lactation

Pregnancy has a paradoxical relation to breast cancer risk. In a young woman pregnancy gives long-term protection against breast cancer, as does parity.⁹ By contrast, pregnancy late in the reproductive age of a woman results in a transient increase in the risk of breast cancer,¹⁰ probably because the hormonal milieu of pregnancy accelerates the clinical presentation of a pre-existing breast tumour.

Lactation offers substantial protection. Analysing forty-seven studies the Collaborative Group¹¹ concluded that breast cancer risk is reduced by 4.3% for every twelve months of breastfeeding and by 7% for each birth. When this formula is applied to the fertility patterns of Western countries, an increase in fertility rates and breastfeeding duration to those seen in developing countries is predicted to halve the lifetime risk of breast cancer.

THE COMBINED PILL AND BREAST CANCER

The notion that the combined pill might promote breast cancer is plausible because breast tissue is a target for female steroidal hormones. The breast's response to oestrogen and progesterone depends on the presence of their corresponding receptors. Breast tumours in perimenopausal and postmenopausal women are more likely to be oestrogen and progesterone receptor positive than breast cancers in women under the age of 40.⁷ Against this idea is the fact that use of the combined pill is highest in women in their 20s (up to 75% exposure¹²), whereas in women over 35 it is low. A major concern is whether there is a latent effect of the pill on breast cancer. The issue is whether a woman using the pill in her 20s may continue to have an increased cancer risk into her perimenopausal and postmenopausal years. Epidemiological research based on well conducted large observational studies has clarified the picture beyond doubt: all studies are consistent in showing no latency.^13,14

The epidemiological evidence

All the epidemiology is based on observational studies. Up to 1996 at least thirty had been published (six prospective cohort studies and twenty-four case-control studies, some using population controls and others using hospital controls). Except for one small case-control study, none of the above showed a significant excess risk of breast cancer in pill users. There was no evidence of an increase in the lifetime risk of breast cancer and there did not seem to be a duration-of-use effect.

The Collaborative Group Study

The Collaborative Group Study,¹³ a major reanalysis of publications worldwide, showed a small excess risk of breast cancer in current and recent combined pill users. The relative risk, 1.24, was barely significant. The risk fell to 1.16 at 1-4 years and disappeared at ≥ 10 years and after discontinuation of the pill. The risk was confined to non-metastatic breast disease. The relative risk for metastatic disease was less than 1. Other findings from this major reanalysis were:

• There was no excess risk in women aged over 45

• There was no synergy with family history of breast cancer

• There was a protective effect against benign breast disease

• There was no duration-of-use effect, no pill-type effect and no effect of age at first use.

The Collaborative Group findings confirmed that the pill was not acting as an initiator of breast cancer and at worst was a promoter of existing disease. The association may even be a surrogate for some other biological or environmental risk factor.

The evidence today

In October 2002 a large case-control study was published from the USA.¹⁴ It included 4575 cases of breast cancer, (compared with the 90 in the UK Oxford Family Planning cohort study). The US workers looked at breast cancer risk in women aged ≥ 35 who were either current or past users of the combined pill. No excess risk was seen in any of the groups studied up to age 64. Moreover, the project confirmed findings from previous studies that duration of use, age at first use, type of pill, type of progestogen and family history of breast cancer did not impact on risk. This is epidemiology at its most robust. It gives the strongest pointer to absence of an association between the combined pill and breast cancer in perimenopausal and postmenopausal women. The questions about the risk in women under the age of 35 remains to be answered, though it is comforting to remember that the majority of cancers in young women are oestrogen receptor negative.

Hormonal contraception and the characteristics of breast cancer

All epidemiology indicates that the small excess risk of breast cancer with hormonal contraception is limited to localized disease. The risk of metastatic disease is slightly reduced in combined pill users. The mechanism behind this differential risk has been a matter of speculation. Pill users are exposed to enhanced surveillance by the health provider, tend to be instructed in breast awareness and probably have greater access to screening opportunities. According to one theory, the pill accelerates the clinical presentation of existing breast tumours, allowing their early detection. This combined with enhanced surveillance leads to diagnosis before occurrence of metastasis. Another idea is that the pill promotes a less aggressive type of breast cancer, and this is supported by the observation that ex-pill users had the same reduced risk of advanced disease as current users. The explanation may lie in the role of cyclin D1,¹⁵ a cell cycle regulator expressed in breast cancer and inhibited by anti-oestrogens. Cyclin D1 tends to be overexpressed in breasts of women on the pill, particularly early users. It is associated with well-differentiated hormone-sensitive tumours.

The young pill user

The Collaborative Group findings of an excess risk of breast cancer of 24% in young pill users translates into a small number of actual cases. A teenager aged 16-19 years has an excess risk of breast cancer of 1 in 20 000 if she uses the pill for 5 years.

Young women who have a first-degree relative with breast cancer tend to be concerned about the impact of the pill on their risk. They can be reassured that, though their background risk is higher, hormonal contraception will have no synergistic/multiplicative effect. Women who have the BRCA1 and 2 gene mutations have an 85% lifetime risk of breast cancer. Such women are also at excess risk of ovarian cancer and colon cancer. The combined pill reduces the risk of ovarian and colon cancer but may increase the risk of breast cancer slightly in women under the age of 35. A careful risk/benefit assessment is required. There is an argument that, since breast cancer can be screened for and treated more successfully than ovarian cancer, benefits of the combined pill may outweigh the risks. We need further evidence on this issue. Pill users over the age of 35, who we now believe have no excess risk of breast cancer, may benefit from the ovarian cancer protection offered by the combined pill.

OTHER HORMONAL CONTRACEPTIVES AND BREAST CANCER

The progestogen-only pill

The Collaborative Group found a non-significant 17% increase in the risk of breast cancer in women taking the progestogen-only pill. ¹³ The evidence was based on a tiny number of cases. Even if the risk is real, it is probably lower than that for the combined pill, and may be negligible.

Injectable progestogens and hormone-releasing devices

Pooled data from several case-controlled studies show a non-significant excess risk of breast cancer with depot medroxyprogesterone. ^13,16 There is no apparent duration-of-use effect and the risk seems restricted to women under the age of 35. There is still some uncertainty as to whether recent use (within 5 years) of depot medroxyprogesterone is associated with a transient increase in the risk of breast cancer. There are no data on subdermal implants and the levonorgestrel-releasing intrauterine system.

CONTRACEPTION FOR WOMEN WITH A PERSONAL HISTORY OF BREAST CANCER

Women who develop breast cancer when young, and survive, will need advice on contraception, the hazards of pregnancy, and later the pros and cons of hormone replacement therapy. The first thing is to assess the patient's fertility potential and aspirations, and then liaise with her oncologist. The general view is that hormonal methods will be contraindicated in the first 2 years after treatment of breast cancer. Barrier methods will be preferred and an intrauterine copper device could be an option after careful counselling. The World Health Organization Medical Eligibility Criteria for Contraceptive Use¹⁷ grade recent breast cancer (within 5 years of diagnosis) as category 4 (hormonal methods contraindicated). For women over 5 years post-diagnosis the combined pill is category 3 (risks outweigh benefits), which means it may be an option as a last resort. Other hormonal contraceptives are classified as follows: injectable progestogens, category 4; subdermal implants, category 4; progestogen-only pill, category 3 for initiation and 4 for continuation; the intrauterine system (Mirena), category 3 for initiation and 4 for continuation.

As regards pregnancy, the Royal College of Obstetricians and Gynaecologists recommends waiting 2-3 years after diagnosis.

CONCLUSION

Although very unlikely to initiate breast cancer, the pill may promote the growth of pre-existing tumours in some women. It does not increase the lifetime risk of breast cancer. Women over 35, whether current or past users, do not show an increased incidence. In women in their 20s the absolute risk of breast cancer is very low, and any excess attributable to the pill is small (around 1 in 10 000). For women as a whole, prolonged or early use seems to make no difference to risk, and there is no synergy with a family history or genetic predisposition. The non-contraceptive benefits of the pill, especially protection against ovarian cancer, need to be considered when a woman is advised on her contraceptive choices.