Ethics Statement

This study was approved by the corresponding ethics committee of each area (Comité ético de investigación clínica de Asturias, Barcelona, Cantabria, Girona, Gipuzkoa, Huelva, León, Madrid, Navarra and Valencia) and informed written consent was obtained from parents. The MCC-Spain study also followed the Declaration of Helsinki and the Spanish Personal Data Protection Act of 1999.

Study design and population

The Multi Case-Control (MCC-Spain) study has been described in detail [20]. Briefly, it is a population-based case-control study of common tumors in Spain; the recruitment includes incident cases of colorectal, breast, gastroesophageal and prostate cancer diagnosed between September 1st, 2008 and December 31st, 2013. Henceforth, we will only refer to breast cancer cases and their controls.

All cases of breast cancer included were incident and pathology confirmed, with no previous diagnosis of breast cancer; they were aged between 20 and 85 years old, and resident within the influence area of the hospital for at least 6 months prior to recruitment in 10 Spanish provinces (Asturias, Barcelona, Cantabria, Girona, Gipuzkoa, Huelva, León, Madrid, Navarra and Valencia). Controls with no prior history of breast cancer were selected from the general population according to age and regional distribution of the cases included in the study. In this paper, 1736 cases of breast cancer in women and their 1895 frequency-matched controls were considered. Response rates were 71% for breast cancer and 72% for controls, with no differences in the main socio-demographic variables among those who participated and those who refused to participate.

Exposure data

Participants were interviewed face-to-face by trained interviewers, using a comprehensive epidemiological questionnaire that collected socio-demographic information, personal and family history of cancer, anthropometric data, smoking habits, alcohol intake, occupation, physical activity, water consumption, reproductive and medical history and medication use, family history, sun exposure, sleep habits, use of hygiene products and cosmetics, signs and symptoms. Comprehensive dietary habits were obtained with the use of a validated food-frequency questionnaire.

Participant’s weight was self-reported, as estimated one year before diagnosis for cases and one year before the interview for controls. Accordingly, body mass index (BMI) was calculated considering self-reported weight, referred to that date, and height. Total fat and vegetable intakes were estimated from the questionnaire using local food composition tables. Similar estimates provided total energy consumption. Physical activity was recorded for the longest occupation and also considering recreational physical exercise.

Detailed information was obtained on past medical conditions and the corresponding medications used. The age at onset, the dates of diagnosis or occurrence and the type of treatment received for each condition was also registered.

Drug use assessment

Drug use was recorded by indication. For each drug, the brand name, dose and duration of exposure were recorded to identify patients with regular drug consumption (“no” and “occasionally” versus “yes”) and the duration of consumption.

The drugs were coded following the Anatomical Therapeutic Chemical Classification System (ATC codes) to define groups with similar mechanisms of action [WHOCC Homepage. WHO Collaborating Centre for Drug Statistics Methodology].

All drugs indicated for the treatment of hypertensive diseases have been considered. The ATC codes included in the present analysis are code C02 (Antihypertensive), C03 (Diuretics), C04 (Peripheral vasodilators), C07 (Beta blocking agents), C08 (Calcium channel blockers) and C09 (Agents acting on the renin-angiotensin system). Results will be presented for each separate group and for specific antihypertensive drugs that had a prevalence of use over 1% in controls.

Statistical Methods

Unconditional binomial logistic regression was used to assess the association between antihypertensive drug use and breast cancer overall and stratifying for menopausal status and BMI (<25/≥25 kg/m²). In order to study the relationship between antihypertensive drug use and different breast cancer subtypes, we applied multinomial logistic regression models; multinomial logistic regression is useful when the outcome is categorical rather than dichotomic; for instance, according to cancer stage, participants were classified in one out of three categories: control / breast cancer stage I-II / breast cancer stage III-IV. Multinomial logistic regression allows to estimate odds ratios for every category (i.e.: breast cancer stage I-II / breast cancer stage III-IV) comparing with the reference category (i.e.: control) [21]. Statistical models were adjusted for the following confounders: age, area of residence, education, BMI 1 year before, active smoking, alcohol intake in the past, family history of breast cancer, age of menarche, age at first full-term birth, parity, menopausal status and hormonal therapy.

Stratified models were developed according to menopausal status, BMI (<25/≥25 kg/m²), clinical stage (I-II / III-IV), ductal (ductal/non ductal), invasive and immunohistochemistry (hormone + receptors with Erbb2 negative, Erbb2 + receptors and triple negative receptors). Results, which are reported only in strata with at least 5 cases or controls using antihypertensive drugs are shown as odds ratios (OR) with 95% confidence intervals (CI). All reported p-values are two-tailed. Statistical analysis was carried out using the package Stata 14/SE (StataCorp, College Station, Tx, US).

Antihypertensive drug consumption and risk of breast cancer according to women’s characteristics

Table 3 and S1 Table–displaying the duration of consumption- show the relationship between the use of antihypertensive drugs and the risk of breast cancer overall- and stratified by menopausal status and BMI. No significant associations were found between breast cancer and any antihypertensive drug for all women combined (Table 3). Users of any antihypertensive drug doubled the risk of developing breast cancer in premenopausal women (OR = 2.15, 95% CI = 1.17 to 3.96).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Relationship between antihypertensive drug consumption and breast cancer according to women’s characteristics.

Category reference no antihypertensive treatment.

https://doi.org/10.1371/journal.pone.0159672.t003

When examining specific classes of antihypertensive drugs, the use of CCBs was associated with a 72% increased risk of breast cancer in the postmenopausal group (OR = 1.72, 95% CI = 1.05 to 2.80) and twice the risk in women with BMI ≥ 25 (OR = 2.05, 95% CI = 1.16 to 3.63); there is moderate confirmation of both associations in women taking CCBs for 5 years or more. Angiotensin II receptor blockers (ARB) were the only group associated with an increased risk of breast cancer in premenopausal women (OR = 4.27, 95% CI = 1.32 to 13.84), but not in postmenopausal women (p for ARB–menopausal status interaction = 0.03). The use of diuretics, beta blockers and angiotensin converting enzyme inhibitors (ACEI) was not associated with breast cancer risk in any women or in different strata.

Antihypertensive drug consumption and risk of breast cancer according to tumor characteristics

Results of the association between the use of antihypertensive drugs and incident breast cancer according to tumor characteristics are shown in Table 4 and in S2 Table for the duration of consumption. Altogether, the use of antihypertensive drugs was associated with a higher risk of triple negative breast cancer (OR = 2.21, 95% CI = 1.37 to 3.56; p for heterogeneity = 0.03); this result was consistently reproduced in women undergoing antihypertensive treatment for more or less than 5 years (S2 Table). Antihypertensive drugs were also associated with more aggressive or worse prognosis cancer: taking antihypertensive drugs increased the risk of developing a tumor in clinical stage III-IV (OR = 1.62; 95% CI = 1.04 to 2.52; p for antihypertensive drug–clinical stage heterogeneity = 0.22), non-ductal tumor (OR = 1.49; 95% CI = 1.00 to 2.24; p for antihypertensive drug–ductal cancer interaction = 0.32), invasive cancer (OR = 1.26, 95% CI = 1.01 to 1.57; p for antihypertensive drug–invasive cancer interaction = 0.02), although results taking into account the treatment length did not reveal a risk pattern consistent with higher risk in women taking antihypertensive drugs for more than 5 years (S2 Table).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 4. Relationship between antihypertensive drug consumption and breast cancer according to tumor characteristic.

Category reference: no antihypertensive treatment.

https://doi.org/10.1371/journal.pone.0159672.t004

Looking at specific classes of antihypertensive drugs, CCBs were even more strongly associated with more aggressive tumors, multiplying by 2.7 the risk of tumors in stage III-IV (OR = 2.70, 95% CI = 1.23 to 5.95), non-ductal cancers (OR = 2.63, 95% CI = 1.27 to 5.43), and Erbb2+ cancer (OR = 2.52, 95% CI: 1.18 to 5.37). CCBs were also associated with invasive tumors (OR = 1.67, 95% CI = 1.04 to 2.70). Similar results were found in women taking CCBs for at least five years.

The use of diuretics, beta blockers, ACEIs or ARBs was not associated with increased risk for specific tumor characteristics.

Other antihypertensive drugs

ARBs were the only antihypertensive group associated with an increased risk of breast cancer in premenopausal women in our results; using ARBs before menopause was, however, scarce; therefore, when stratifying by length of consumption the results were non-significant, although the odds ratios scaled from 3.48 for less than 5 year users to 6.64 for 5 or more year users. We have not found other papers analyzing the relationship between ARBs and breast cancer in premenopausal women; this together with the small number of premenopausal women taking ARBs make the interpretation of this result highly speculative. Bhashkaran et al (2012) found an increased risk of breast cancer associated with short-term exposure to ARBs; they suggested that such an association could not be causal but the result of confounding by indication: according to them, some early symptoms of breast cancer could induce ARBs to be indicated instead of ACEIs in hypertensive patients, leading, therefore, to a spurious ARB–breast cancer association [33]. We have no data for exploring such an explanation in our study.

Differences in study design and population characteristics may explain the conflicting results reported on the antihypertensive drugs–breast cancer association. Inferring is difficult because of the small sample size in some studies, differences in the populations evaluated and designs (cohort of patients, general population cohorts and case-control studies). In many cases, randomized trials cannot identify long-term adverse effects of medication because they are usually conducted for relatively shorter periods (i.e.: 5 years or less) [34]. Subsequent long-term monitoring of drugs through observational studies may overcome this limitation and provide new information in this regard. Some observational studies have analyzed the antihypertensive drugs-breast cancer relationship by working with administrative data, which were not designed for this objective. That kind of design does not allow an adequate adjustment for confounding factors. Although some studies linked the use of CCB and cancer in the 90s, antihypertensive consumption was different to today’s. Nowadays, more therapeutic options are available; the use of antihypertensive drugs is characterized by the appearance of new fixed-dose combinations of two active antihypertensive drugs and by the introduction of new drug treatments belonging to the group of ARBs, beta blockers or CCBs [1].

The present study has some limitations. First, recall bias should be considered as in any case-control study. Drug consumption was obtained using a standardized questionnaire in face-to-face interviews where examiners were blinded to the case-control status. If a non-differential recall bias was produced, then the odds ratios should be downward biased and the positive associations we have found for CCB, ARBs or ACEIs would actually be even stronger than reported here. If a differential recall bias were responsible for these associations, breast cancer cases would also have over-declared (or controls under-declared) their consumption of other hypertensive drugs; however, no association has been found between breast cancer and diuretics or beta-blockers, which makes a differential recall bias less probable. Second, some strata in our study have small numbers of exposed cases or controls; this could produce unstable estimates. This limitation should be especially considered regarding ARBs, as their relationship with breast cancer is confined to premenopausal women, a relatively small subgroup whose exposure to hypertensive drugs could not be too long. Third, antihypertensive consumption was reported by indication; therefore, most people taking antihypertensive drugs have hypertension. Thus, we cannot distinguish using antihypertensive drugs from having hypertension. Lastly, many comparisons have been made, raising the probability of finding some spurious results However, the consistency of some results in different subgroups, such as those for CCBs, supports the existence of a real excess risk associated with their use.

In summary, we report that consumption of antihypertensive medications -as a global group- was associated with an increased odds of breast cancer in premenopausal women, and the use of CCBs in particular was associated with an increased odds of breast cancer in postmenopausal women and those with BMI higher than 25 Kg/m². As people with hypertension are expected to take antihypertensive drugs for many years, their relative effect on breast cancer should be taken into account when choosing the antihypertensive to be prescribed.