Risk of Breast Cancer With Long-term Use of Calcium Channel Blockers or Angiotensin-Converting Enzyme Inhibitors Among Older Women

Marsha A. Raebel; Chan Zeng; T. Craig Cheetham; David H. Smith; Heather Spencer Feigelson; Nikki M. Carroll; Kristin Goddard; Heather M. Tavel; Denise M. Boudreau; Susan Shetterly; Stanley Xu

Disclosures

Am J Epidemiol. 2017;185(4):264-273. 

In This Article

Discussion

In this large cohort of older US women with hypertension, in comparison with taking a CCB for 1–<2 years, taking a CCB for 2–<12 years was not associated with increased risk of invasive breast cancer. Taking an ACEi for 2–<12 years conferred a protective association against breast cancer as compared with taking an ACEi for 1–<2 years.

These findings were consistent regardless of whether or not the cumulative exposure to CCBs or ACEis included stockpiled medication. When risks of breast cancer were compared between women taking CCBs and those taking ACEis in a subcohort matched on length of follow-up, the duration association of ACEi use differed from the duration association of CCB use. It appears that the similarity of follow-up times strengthened the protective association of ACEi use.

The protective association with ACEi use was observed within 2–<3 years of initiation (Table 2). Point estimates declined, plateaued, and declined again across years 3–<12. After year 7–<8, the 95% confidence intervals were wider. We could not determine whether there was a further risk reduction after year 8.

Our finding that CCB use does not increase breast cancer risk is consistent with most cohort,[1,4,6,31,32] case-control,[7,8,33,34] and nested case-control[35] studies. Our results differ from those of 2 other US studies, a cohort study by Saltzman et al.[11] and a case-control study by Li et al..[9] They also differ from those of a case-control study conducted by Leung et al.[36] in Taiwan.

Saltzman et al. found that, compared with women with no use of antihypertensive medication, women with recent CCB use had a 1.6-fold increased risk of breast cancer (95% CI: 1.0, 2.5).[11] Differences in study design and analysis between their study and ours that could contribute to different findings include the fact that they did not have a new-user cohort and probably had incomplete CCB data because use was determined from prescription label review. Further, they collected comorbidity data from self-administered questionnaires (potential recall bias). Finally, they did not consider concomitant use of antihypertensive agents or mammography.

In the Li et al. study, CCB use for ≥10 years was associated with a 2.4-fold higher risk of ductal breast cancer (95% CI: 1.2, 4.9) and a 2.6-fold higher risk of lobular breast cancer (95% CI: 1.3, 5.3).[9] Li et al. reduced recall bias by focusing on current CCB use and had interviewers record use from participants' prescription bottles. However, some exposure misclassification probably remained. They did not consider concomitant use of antihypertensive agents.

Leung et al. found a 16% increase in breast cancer risk (95% CI: 1.06, 1.28) among women in the highest quartile of cumulative CCB dose and increasing risk with increasing duration of exposure (test for trend: P = 0.006).[36] Leung determined CCB dispensings from the Taiwan National Health Insurance Research Database. That database does not include information on body mass index, a breast cancer risk factor. They did not consider concomitant use of antihypertensive agents.

Most published studies of ACEi use and breast cancer have found either no increased risk or a nonsignificant risk reduction,[1,4,7–10,36,37] including the Li et al. study discussed above.[9] Li et al. found ACEi use for ≥10 years to be associated with an odds ratio of 0.7 (95% CI: 0.5, 1.2) for invasive ductal carcinoma and an odds ratio of 0.6 (95% CI: 0.4, 1.0) for invasive lobular carcinoma.[9] As we mentioned above, their study had potential exposure misclassification and did not account for use of multiple antihypertensive agents. In addition, the size of our study population exceeded that in the Li et al. study, potentially contributing to our significant finding.

Our findings reflect differences in ACEi and CCB prescribing patterns. For example, ACEis may be preferred in women with diabetes for their kidney-protective effects (35.8% of the ACEi group had diabetes vs. 20.9% of the CCB group). Dual antihypertensive therapy with a thiazide-type diuretic is common (thiazide-type diuretics were used by 68.3% of the ACEi group and 53.6% of the CCB group). Use of CCBs as first-line antihypertensive agents has declined recently and may account for the low proportion of women (17.9%) initiating use of a CCB and the high proportion (36.3%) switching to an ACEi.

Our findings that long-term use of CCBs was not associated with increased risk of breast cancer and that long-term use of ACEis may be associated with reduced risk of breast cancer have public health and treatment implications. The prevalence of hypertension among women aged ≥55 years ranges from 52% (at age 55–64 years) to 80% (at age ≥75 years),[38] and most hypertensive women are prescribed medication. Our findings reassure prescribers and women with hypertension about the association between long-term use of CCBs or ACEis and breast cancer.

The proposed mechanism by which CCBs increase breast cancer risk was based on cell-line studies demonstrating that blockade of calcium channels inhibited apoptosis, enabling division of damaged cells with malignant potential.[39] Subsequent evaluation did not support this hypothesis.[40]

The proposed mechanism by which ACEis reduce breast cancer risk over time is related to angiotensin II. Angiotensin II stimulates neovascularization and acts as a growth factor.[10] ACEi inhibits angiogenesis and the growth of induced cancer in rats.[10] Given that angiotensin II production decreases soon after ACEi initiation, our findings that the protective association was observed relatively early and that there might not be additional risk reduction after years of therapy is plausible. Further research would be needed to clarify this.

In addition to the fact that our cohort included thousands of women with years of CCB or ACEi use, another strength of our study is that we determined CCB or ACEi use from prescription dispensings, a more accurate method than self-reporting.[1,8,9,11,33,34] Other strengths are that we included women with diagnosed hypertension (reducing potential confounding by indication) and identified confounders and cancers using longitudinal databases. Finally, to ensure that immortal time bias was not a concern, we used an exposure-defined cohort, setting the index date to the date of the first dispensation of a CCB or ACEi. Even if the date of a woman's qualifying prescription fell between her first and second hypertension diagnoses, the person-time between diagnoses was not event-free. The definition of ACEi or CCB initiation was the same for all women, and breast cancer could occur at any time after initiating use of the ACEi or CCB.[41–43]

Our study also had limitations. Residual confounding is a concern in any observational research. To reduce the potential for residual confounding, we adjusted for known confounders. However, some possible confounders, such as age at menopause, could not be controlled for because information on them was not available in the electronic data. We believe misclassification of menopausal status was very low, because ≥95% of women reach menopause by age 55 years.[15,44] Our definition of "new user" required no CCB or ACEi use in the prior year, and a few women could have had more remote exposure. However, missed use prior to cohort entry would have resulted in underclassification of total duration. It is also feasible that a few women could have used β-blockers, angiotensin receptor blockers, or diuretics prior to the look-back period only. Similar to previous studies of cancer risk and medication use, we did not consider the competing risk of death with longer duration of use, nor did we consider the duration of hypertension prior to ACEi or CCB initiation. Finally, including only women with diagnosed hypertension reduced the generalizability of our findings.

We conclude that CCB use for up to 12 years does not increase the risk of developing invasive breast cancer in older women with hypertension. We further conclude that treatment with ACEis for up to 12 years may confer protection against the development of invasive breast cancer. These findings can assist prescribers when they are counseling older women with hypertension on the long-term use of CCBs or ACEis.

processing....