Risk of cancer and the oral contraceptive pill

BMJ 2007; 335 doi: http://dx.doi.org/10.1136/bmj.39336.503067.BE (Published 27 September 2007) Cite this as: BMJ 2007;335:621
  1. Olav Meirik, senior research associate1,
  2. Timothy M M Farley, scientist2
  1. 1Instituto Chileno de Medicina Reproductiva (ICMER), Jose Ramon Gutierrez 295, Santiago, Chile
  2. 2Department of Reproductive Health and Research, World Health Organization, 1211 Geneva 27, Switzerland
  1. farleyt{at}who.int

    Long term follow-up of women in the UK shows no increased risk

    In the preface to the first comprehensive publication from the Royal College of General Practitioners' oral contraception study Sir Richard Doll wrote, “Final judgement [on the safety of the pill] must still await the passage of time, when observations can be made of women who have used the pill for 10 or 20 years.”1 Thirty years later, in this week's BMJ, Hannaford and colleagues2 report incidence rates of cancer in relation to use of the pill among women in the study cohort.

    Between 1968 and 1969, 45 950 women in the United Kingdom were enrolled in the study, and they were followed for a mean of 24 years. Full assessment of the risk of cancer needs a long follow-up as effects of the pill may persist many years after its use has been stopped. Incidence rates of cancer in women who ever used the pill were compared with rates in women who never used the pill. On balance, no higher risk of cancer was found in pill users. Risks were significantly lower for cancer of the colon or rectum, uterine body, or ovaries; the main gynaecological cancers combined (uterine body, ovaries, cervix); and for any diagnosis of cancer. The incidence of breast cancer was similar in pill users and never users.

    These data came from six monthly reports from the women's general practitioners until 1996, and from linkage of the 35 050 women still in the study in the mid-1970s to National Health Service central registries. These provided cancer diagnoses until 2004 to supplement those reported by general practitioners. The follow-up covered two thirds of the woman years that would have accumulated if all 45 950 women had been followed from 1968 or 1969 to 2004.

    Hannaford and colleagues also report analyses restricted to follow-up by general practitioners until 1996, which allow incidence rates to be calculated according to duration of pill use and time since stopping use of the pill. The comparisons between ever users and never users were largely similar for the two sources of data. After adjustment for age, parity, smoking, social class, and use of hormone replacement therapy, the relative risks of cancers of the ovary and uterine body in ever users compared with never users were below unity for all durations of pill use (≤48, 49-96, and ≥97 months). The opposite was found for cancers of the cervix and the brain or pituitary—relative risks increased progressively with longer use. The patterns of risks by time since stopping the pill were largely reassuring, although some excess risk of cervical cancer persisted 10-15 years after stopping, and risk of brain or pituitary cancer persisted 20 or more years after stopping. Some individual risk estimates are significantly different from 1.0; for example, risk of breast cancer was increased 15-20 years after stopping, yet was significantly reduced 20 or more years after stopping. Considering the many comparisons included in the paper, individual risk estimates must be interpreted with caution.

    The data from this and other studies indicate that pill use prevents or postpones ovarian and endometrial cancers3 4 but probably accelerates development of cervical cancer caused by chronic infection with oncogenic human papillomavirus.5 Fortunately, preinvasive cervical cancer can be detected by cervical cytology and treated. Regular cervical cytology screening remains an important element of quality health care, particularly for women who use the pill.

    The finding that pill use increases the risk of brain or pituitary cancers may result from prescription bias—menstrual disturbances are often regulated by the pill, and such disturbances may be an early symptom of pituitary disease. One study of pituitary prolactinomas and pill use found odds ratios of 7.7 (95% confidence interval 3.5 to 17.0) in women prescribed the pill for treatment of menstrual irregularities and 1.3 (0.7 to 2.6) in women prescribed the pill for contraception.6

    Considering that the Royal College of General Practitioners' study enrolled women almost 40 years ago, the age distribution of pill users is remarkably similar to current patterns of use. At enrolment, 61% of pill users were under 30 years, and the age group 20-24 years had the largest number of women.1 A UK survey in 2005-6 showed that 52% of pill users were under 30 and the highest prevalence of pill use was among 20-24 year olds.7

    Pills used in the late 1960s and 1970s contained higher dosages of progestogens and oestrogen (ethinylestradiol) than the currently widely used 30 µg ethinylestradiol combined pills. Pill users in the study would have started with higher dose pills, with a progressive switch to the lower dose formulations used today. Limited data suggest that the reduced risks of ovarian and endometrial cancers are maintained with lower dose pills, so that the overall balance of cancer risks can be expected to apply to today's pill users.

    The results of this unique long term study agree with findings from the Oxford Family Planning Association's cohort study and modelling studies.3 4 The results show that—in a developed country with an effective cervical cancer screening programme—the pill is a safe contraceptive method with respect to cancer. In some developing countries—with inadequate cervical cancer screening and healthcare services, and high cervical cancer rates—the balance of cancer risk is probably less favourable.8 However, in such settings, contraceptive benefits must be weighed against the risk of cervical cancer, and the balance would tilt in favour of the pill because of the high morbidity and mortality associated with unplanned pregnancies.


    • This article was posted on bmj.com on 12 September 2007: http://bmj.com/cgi/doi/10.1136/bmj.39336.503067.BE

    • Competing interests: OM has received consultancy fees from Wyeth Pharmaceuticals. TMMF is a staff member of the World Health Organization. He alone is responsible for the views expressed in this publication and they do not necessarily represent the decisions or the stated policy of the WHO.

    • Provenance and peer review: Commissioned; not externally peer reviewed.