Globally, hormonal contraceptives (HCs) are used by approximately 13% of the female
population between 15 and 49 years old, which, in absolute numbers, represents 140
million women.[1] In Brazil, where the rates of tubal ligation are 29%, HCs are used by approximately
30% of women of reproductive age.[2]
The roles of the hormones contained in HCs have long been discussed. Estrogens appear
to function as promoters of mammary tumors; however, their role as initiators of carcinogenesis
is controversial, as is the complex role of progestins.[3] Studies that address the association between breast cancer and HCs are usually restricted
to combined oral contraceptives (COCs). Even though older studies have shown greater
risk of breast cancer among women on the pill,[4] since 2002, the evidence has not demonstrated this association.[5]
[6]
[7]
[8]
In December 2017, the New England Journal of Medicine published a study by Mørch et
al[9] in which they presented the results of a 10.9-year prospective cohort study involving
1.8 million women in Denmark who were on HCs. The study investigated the association
between HCs and the risk of breast cancer for this population, when compared with
women who did not use HCs.
During the follow-up period, 11,517 cases of breast cancer occurred. The relative
risk of breast cancer among recent users of any type of HC was of 1.20 (95% confidence
interval [95%CI]: 1.14-1.26). Among the different combination (estrogen-progestin)
contraceptives, the relative risk varied between 1.0 and 1.6. Higher risk of breast
cancer was also observed among users of the levonorgestrel intrauterine system (relative
risk [RR]: 1.21; 95%CI: 1.11-1.33). Furthermore, the risk increased with longer durations
of HC use.
In general, critical analyses of high-impact articles tend to focus on possible biases
that, for the most part, are already addressed by the authors themselves. This includes
the difficulty in controlling confounding factors, such as age of menarche, breastfeeding,
alcohol consumption, physical activity, and body mass index. Moreover, the Danish
study also showed that the risk of breast cancer usually disappeared rapidly after
the interruption in the use of these contraceptives.
Indeed, there are no perfect studies. However, the quality of the study published
by the Danish group is undeniable. Thus, the key question with regard to this publication
would be better framed as the impact that its dissemination might have on physicians
and, ultimately, patients. Essentially, the study assessed the risk of women presenting
the disease they fear the most. The term “relative risk” was used to address the topic,
as is common in the medical literature. In the introduction, the authors reported
that the risk of cancer among HC users could increase by 20% to 30%. This statement
was absolutely correct, based on studies that have shown relative risks ranging between
1.2 and 1.3. However, when the subject is addressed using risk percentages, which
are typically expressed in terms of relative risk, the scale of such risks may be
overrated, especially among the non-medical audience. When presented in terms of incidence
or prevalence, risks of disease are easier to understand.
Referring to absolute risk is more precise. Thus, according to the study by Mørch
et al,[9] the absolute risk of breast cancer associated with any HC was 13 per 100,000 women-years,
which corresponds to 1 extra case of breast cancer for every 7,690 HC users in 1 year.
Clinical judgements after having analyzed these data led medical societies, such as
the Brazilian Federation of Gynecology and Obstetrics Societies (Febrasgo, in the
Portuguese acronym), to clarify for women the safety of using hormonal contraceptive
methods.[10]
Generally speaking, the perception of risks among patients, and even physicians, varies
depending on how these risks are presented.[11] Thus, there is a pressing need for physicians, as primary communicators, to analyze
the data, translate them, and provide patients with clear, accurate explanations,
avoiding controversy and inappropriate conduct. Communicating research results in
terms of absolute risk seems to be the most balanced approach.
