Statin use and risk of breast cancer
Findings that statins inhibited the proliferation of breast cancer cells in vitro and in rodents have raised interest in whether the use of statins might decrease a woman's risk of developing breast cancer. We analyzed data from a population-based case-control study to evaluate the association between the use of statins and breast cancer risk.
Cases of incident invasive breast cancer in women 50 years of age or older and diagnosed from 1995–2001 were identified from population-based cancer registries in Wisconsin, Massachusetts, and New Hampshire. Controls were randomly selected, within each state, from lists of licensed drivers and Medicare beneficiaries. Information on the use of statins and other breast cancer risk factors was ascertained from structured telephone interviews.
Overall, breast cancer cases were not more likely than controls to have ever used statins. Ever use of lipophilic statins as a group (simvastatin, lovastatin, and fluvastatin) and ever use of the hydrophilic statin pravastatin were also not associated with breast cancer risk. Ever use of fluvastatin was associated with a decreased risk of breast cancer (odds ratio [OR], 0.5; 95% confidence interval, 0.3–0.8) but the magnitude of the ORs did not vary across categories of duration of use.
The use of statins overall was not associated with breast cancer risk. Cancer 2008. © 2007 American Cancer Society.
Statin drugs (3-hydroxy-3 methylglutaryl coenzyme A reductase inhibitors) lower blood cholesterol levels and reduce mortality from cardiovascular disease.1 The first statin approved for use in the US was lovastatin in 1987 and several other statin drugs are now available to treat hypercholesterolemia.2 Findings from studies of breast cancer cell lines3 and in rodents4 that showed that statins inhibited breast cancer cell proliferation have fueled interest in whether the use of these drugs might reduce a woman's risk of developing breast cancer. However, the evidence is not consistent5; for example, at relatively high doses statins were found to increase the risk of several cancers in rodents.6
A meta-analysis of 7 randomized trials, in which the primary outcomes were cardiovascular disease endpoints, and 9 observational studies found no association between the use of statins and occurrence of breast cancer. The randomized trial and observational studies summary odds ratios (ORs) were 1.0 (95% confidence interval [CI], 0.8–1.3) and 1.0 (95% CI, 0.9–1.2), respectively.7 Despite these findings there have been recent calls for randomized trials of statin therapy in women to evaluate whether statin use reduces the risk of breast cancer.3, 4, 8–11 Supporters argue that the lengths of follow-up in previous randomized trials were too short to evaluate breast cancer risk. In addition, specific preparations, or their classes, may have different effects on breast cancer risk.12, 13 For example, statins can be classified based on their solubility in octanol (lipophilicity) or water (hydrophilicity).14, 15 Lipophilic statins can penetrate biological membranes, whereas the hydrophilic statin, pravastatin, cannot.12 Uptake of lipophilic statins by a malignant cell may inhibit its growth because synthesis of malvalonate, a precursor of cholesterol that is required for various cell functions, is reduced.12, 13 However, few studies have evaluated breast cancer risk by specific preparations or classes of these medications.16, 17
In our large multicenter population-based case-control study we collected data on numerous breast cancer risk factors. In this analysis we evaluated the risk of breast cancer associated with the use of statins overall and with the use of specific types of statins.
MATERIALS AND METHODS
Cases were women with a first primary invasive breast cancer diagnosis identified from population-based cancer registries in Wisconsin, Massachusetts, and New Hampshire according to protocols approved by the Institutional Review Boards at each site. Eligible cases included women diagnosed between January 1995 and May 2001 who resided in Wisconsin, Massachusetts (excluding metropolitan Boston), or New Hampshire, aged 20–69 years at diagnosis, had a listed telephone number, and verified by self-report to have a driver's license if they were younger than 65 years of age. After contacting the personal physician of each eligible case subject (n = 8066), 6421 case subjects (80%) were successfully interviewed. In this analysis we restricted the sample to women 50 years of age or older at the reference age (age at diagnosis of the cases and comparable age in the controls) because statin use among women younger than 50 years of age is relatively uncommon.18 Thus, 2242 cases with a reference age younger than 50 years were excluded and 4179 cases remained.
Within each state, controls frequency matched to cases within 5-year age strata were randomly selected from lists of licensed drivers if younger than 65 years of age, or a roster of Medicare beneficiaries compiled by the Centers for Medicare & Medicaid Services (formerly the Health Care Financing Administration) if 65 years of age and older. To be eligible as a control a woman must have had a listed telephone number, no personal history of breast cancer and, if younger than 65 years of age, possess a self-reported driver's license. Among 10,160 eligible control subjects, 7673 (76%) were successfully interviewed. Among these subjects 2690 women with a reference age younger than 50 years were excluded and 4983 controls remained.
Information on the use of statins was ascertained in a structured telephone interview. Women were asked about their use of cholesterol-lowering medications. Those who had ever used a cholesterol-lowering medication were asked to report further information on each episode of use. Details on the type, duration, and timing of each statin used were ascertained. Specifically, women were asked whether they had ever taken a cholesterol-lowering medication; if they responded affirmatively they were asked to name each cholesterol-lowering medication they used and the date it was first and last used. The interview also elicited information on known and suspected breast cancer risk factors including reproductive and menstrual history, alcohol consumption, physical activity, height, weight, exogenous hormone use, family history of breast cancer, screening mammography history, and demographic characteristics.
A sample of cases and controls were reinterviewed to estimate the reliability of our questionnaire (n = 292). Among these, 203 (n = 88 cases and n = 115 controls) were 50 years of age or older. Six cases and 6 controls were excluded because they had missing exposure data either in the original questionnaire or in the reliability questionnaire. Thus, the total sample size for the kappa statistic was n = 191. Only 9 participants reported ever using statins in the reliability questionnaire (4.7%). The kappa for ever versus never use of statins (defined below) was 0.65 (95% CI, 0.39–0.91).
The statins considered in this analysis included the most commonly prescribed preparations approved by the Food and Drug Administration (FDA) for use from 1995 through 2000: simvastatin, lovastatin, fluvastatin, and pravastatin. Statins were considered individually and were grouped according to whether they were lipophilic (simvastatin, lovastatin, and fluvastatin) or hydrophilic (pravastatin).14, 15 A woman who used more than 1 statin was considered a user of each preparation in separate analyses. Never users were defined as never users of any statin and ever users were defined as ever users of any statin for at least 6 months. Former users were defined as ever users of a statin for at least 6 months who last used the medication before the year before the reference age, and current users were defined as ever users of statins for at least 6 months who used the medication in the year before the reference age. Ever use of any nonstatin cholesterol-lowering drug was defined as ever use of clofibrate, colestipol hydrochloride, gemfribrozil, probucol, niacin, or cholestyramine before the reference age. All exposures for cases and controls were limited to those that occurred before the reference age.
Among the 4179 cases and 4983 controls, data were missing on ever use of cholesterol-lowering medication for 4.9% of cases (n = 205) and 2.2% of controls (n = 109). Among the 3974 cases and 4874 controls with known values for ever use of cholesterol-lowering drugs, 1.3% of cases (n = 51) and 1.4% of controls (n = 66) were missing data for the type or duration of use. Ever users of atorvastatin, which was approved by the FDA for use in 1997, were excluded from all analyses (n = 64 cases and n = 47 controls) because the prevalence of the use of atorvastatin increased rapidly after 1997 among the controls, whereas the use of the other statins was relatively stable during this period. After exclusion of ever users of atorvastatin the final sample size included 3859 cases and 4761 controls.
Unconditional logistic regression was used to estimate ORs and Wald 95% CIs for the risk of breast cancer associated with the use of statins.19 Two estimates of the ORs are presented. One estimate is adjusted for the variables that were used for frequency matching (reference age, state of residence, and reference year) and the other estimate is also adjusted for factors considered a priori to be related to the incidence of breast cancer, including first-degree family history of breast cancer, menopausal status/age at menopause, parity/age at first birth, body mass index, recency of postmenopausal hormone use, education, and history of screening mammogram in the 5 years before the reference age. The categorization of these variables is shown in Table 1. Women with a missing value for at least 1 of the variables included in a regression model were excluded from that regression model. An interaction between ever/never use of statins and age 65 years and older/younger than 65 years was tested using the likelihood ratio (LR) statistic by comparing the maximum log likelihood of the model with and without the interaction term.19
Table 1. Breast Cancer Cases and Controls by Select Characteristics
| Less than high school||312||8.1||449||9.4|
| High school graduate||1652||42.9||2088||43.9|
| Some college||930||24.1||1197||25.2|
| College graduate||958||24.9||1020||21.5|
|Age at menarche|
| < 13 y||1747||45.7||2001||42.5|
| 13- < 14 y||1094||28.6||1342||28.5|
| ≥14 y||981||25.7||1367||29.0|
|Age at first birth|
| < 20 y||621||18.4||905||21.0|
| 20–24 y||1753||51.8||2317||53.7|
| 25–29 y||719||21.2||851||19.7|
| ≥30 y||291||8.6||245||5.7|
|Age at menopause†|
| < 45 y||654||21.2||999||26.3|
| 45–49 y||686||22.2||854||22.5|
| 50–54 y||1063||34.4||1232||32.4|
| ≥55 y||382||12.4||405||10.7|
|Recency of postmenopausal|
|First degree family history of breast|
|Body mass index (kg/m2)|
| < 25||1647||43.1||2075||44.0|
| Less than annual||866||22.6||1515||32.0|
| At least annual||2534||66.2||2747||58.1|
| New Hampshire||378||9.8||570||12.0|
Compared with controls, cases were more likely to have lower parity, older age at first birth, a first degree family history of breast cancer, ever used postmenopausal hormones, and reported at least annual mammograms in the 5 years before the reference age (Table 1).
Overall, 7.0% of women ever used a statin (271 cases and 336 controls) (Table 1). For cases and controls combined, lovastatin (2.8%) and simvastatin (2.4%) were the most commonly used, followed by pravastatin (1.6%), and fluvastatin (1.0%) (Table 2). The mean cumulative duration of statin use was slightly greater for cases than controls, 4.9 years (median = 4.0 years) and 4.5 years (median = 3.0 years), respectively.
Table 2. Associations Between Breast Cancer Risk and Statin Use
|Ever use of any statin for < 6 mo||9||0.2||10||0.2||1.1||0.4–2.7||1.0||0.4–2.5|
|Ever use of any statin for ≥6 mo||271||7.0||336||7.1||1.0||0.8–1.2||1.0||0.8–1.2|
| Former use||31||0.8||39||0.8||0.9||0.7–1.1||1.0||0.6–1.6|
| Current use||240||6.2||297||6.2||1.1||0.9–1.4||1.0||0.8–1.2|
| >0.5 to <5 y||126||3.3||168||3.5||0.9||0.7–1.1||0.9||0.7–1.1|
| ≥ 5 to < 10 y||89||2.3||92||1.9||1.2||0.9–1.7||1.2||0.9–1.7|
| ≥10 y||25||0.6||37||0.8||0.8||0.5–1.4||0.8||0.5–1.4|
|Any type of lipophilic statin|
| Ever use for ≥6 mo||214||5.5||276||5.8||0.9||0.8–1.1||1.0||0.8–1.2|
| Former use||30||0.8||38||0.8||1.0||0.6–1.5||1.0||0.6–1.7|
| Current use||184||4.8||238||5.0||0.9||0.8–1.2||1.0||0.8–1.2|
| >0.5 to <5 y||97||2.5||137||2.9||0.9||0.7–1.1||0.9||0.7–1.1|
| ≥5 y||87||2.3||101||2.1||1.1||0.8–1.4||1.1||0.8–1.4|
| Ever use for ≥6 mo||95||2.5||113||2.4||1.0||0.8–1.3||1.0||0.7–1.3|
| Former use||10||0.3||6||0.1||2.0||0.7–5.5||2.1||0.7–6.2|
| Current use||85||2.2||107||2.2||0.9||0.7–1.3||0.9||0.7–1.3|
| >0.5 to <5 y||57||1.5||73||1.5||0.9||0.7–1.3||0.9||0.7–1.4|
| ≥5 y||28||0.7||34||0.7||1.0||0.6–1.6||0.9||0.5–1.5|
| Ever use for ≥6 mo||110||2.9||128||2.7||1.1||0.8–1.4||1.2||0.9–1.5|
| Former use||36||0.9||42||0.9||1.1||0.7–1.7||1.1||0.7–1.8|
| Current use||74||1.9||86||1.8||1.1||0.8–1.5||1.2||0.8–1.6|
| >0.5 to <5 y||28||0.7||36||0.8||1.0||0.6–1.6||1.1||0.6–1.8|
| ≥5 y||46||1.2||50||1.1||1.2||0.8–1.8||1.2||0.8–1.9|
| Ever use for ≥6 mo||25||0.6||57||1.2||0.5||0.3–0.8||0.5||0.3–0.8|
| Former use||0||0.0||5||0.1||‡|| ||‡|| |
| Current use||25||0.6||52||1.1||0.6||0.3–0.9||0.6||0.3–0.9|
| >0.5 to <5 y||18||0.5||41||0.9||0.5||0.3–0.9||0.5||0.3–0.9|
| ≥5 y||7||0.2||11||0.2||0.7||0.3–1.9||1.1||0.4–3.3|
|Hydrophilic statin (Pravastatin)|
| Ever use for ≥6 mo||73||1.9||69||1.4||1.3||0.9–1.8||1.2||0.9–1.8|
| Former use||10||0.3||8||0.2||1.5||0.6–4.0||1.5||0.6–4.1|
| Current use||63||1.6||61||1.3||1.3||0.9–1.8||1.3||0.8–1.7|
| >0.5 to <5 y||40||1.0||37||0.8||1.3||0.8–2.1||1.2||0.7–1.9|
| ≥5 y||23||0.6||24||0.5||1.2||0.7–2.2||1.2||0.7–2.1|
Overall, ever use of statins was not associated with breast cancer risk and the magnitude of the ORs was similar with increasing duration of use among current users compared with never users (Table 2). As a group, ever use of lipophilic statins was not associated with breast cancer risk (OR, 1.0; 95% CI, 0.8–1.2). Ever use of the lipophilic statin fluvastatin was associated with a reduced risk of breast cancer (OR, 0.5; 95% CI, 0.3–0.8). However, the reduction in risk was statistically significant only among users of less than 5 years (OR, 0.5; 95% CI, 0.3–0.9). Few women (n = 7 cases and n = 11 controls) used fluvastatin for more than 5 years (OR, 1.1; 95% CI, 0.4–3.3). The use of other lipophilic statins (simvastatin and lovastatin) was not associated with breast cancer risk regardless of duration or recency (current or past) of use. The use of the hydrophilic statin pravastatin (OR, 1.2; 95% CI, 0.9–1.8) was not associated with risk and the magnitude of the ORs did not vary with recency or duration of use.
To evaluate breast cancer risk associated with the use of statins among women with presumably similar indications for the use of statins, we repeated our analyses restricted to women who ever used any cholesterol-lowering drug (n = 335 cases and n = 389 controls). Women that ever used any statin were at reduced breast cancer risk when compared with users of other cholesterol-lowering medications (OR, 0.6; 95% CI, 0.4–0.9), with no evidence of further risk reduction with increasing duration. Similar results were obtained regardless of statin type (data not shown). However, the reference group for this analysis, ever users of cholesterol-lowering medications who never used statins, included only 64 cases and 53 controls.
Breast cancer risk associated with ever use of statins compared with never use was similar among women 65 years of age and older compared with women younger than 65 years of age (data not shown) (P-value for interaction, .25).
In this multicenter population-based case-control study the use of statins overall, and of lipophilic statins as a group, was not associated with breast cancer risk, regardless of recency or duration. For specific lipophilic statins, the use of simvastatin or lovastatin was not associated with risk, whereas the use of fluvastatin was associated with a reduced risk of breast cancer. We observed no association with pravastatin, a hydrophilic agent.
Most previous studies are consistent with our finding of no association for ever use of statins, regardless of duration, and breast cancer risk.16, 17, 20–27 However, in a prospective cohort study by Cauley et al.28 a reduced breast cancer risk was observed for ever users of statins compared with nonusers among women age 65 years and older. This relation was not evaluated by duration of statin use,28 but most studies show no trend in breast cancer risk related to increasing duration of ever use of statins.16, 20, 21, 23, 25 However, in 1 case-control study there was a suggestion of a reduced risk associated with overall use of statins for longer than 5 years, whereas no association was observed with use for 5 years or less.17 In a retrospective cohort study an increased breast cancer risk was reported for overall use of statins for less than 6 months and a reduced risk with use for longer than 4 years.22
We evaluated statins in groups as well as singly. Similarly, a previous prospective cohort study compared breast cancer risk in nonusers of statins to that in 2 categories of statin users.16 They observed a reduced risk associated with the use of simvastatin, lovastatin, or fluvastatin (hazard ratio [HR] = 0.82; 95% CI = 0.70 to 0.97) and no association with the use of atorvastatin or pravastatin, after adjustment for breast cancer risk factors and mammogram use.
For ever use of individual statins, 1 randomized trial of coronary heart disease endpoints observed a 12-fold increased risk of breast cancer in women assigned to the pravastatin group (12 cases in 291 women) compared with women in the placebo group (1 case in 291 women).29 Other randomized trials of pravastatin with similar or longer durations of follow-up and larger numbers of female participants did not find an association of pravastatin with breast cancer risk.30–32 Two case-control16, 17 and 1 prospective cohort study25 also showed no association between use of pravastatin and breast cancer risk. For the use of other individual statins, no association was observed between breast cancer risk and use of lovastatin,16, 17, 25, 33 simvastatin,16, 17, 25, 34, 35 fluvastatin,16, 17, 25 or atorvastatin.16, 17, 25 Except for 1 other study16 the present study had the largest number of invasive breast cancer cases.
We relied on self-report of past statin use. However, a validation study among women aged 65–79 years that compared self-reported medication use to pharmacy records showed that women recall use of these medications relatively well.36 The specificity of recall of statin use for the 6 months before the reference date did not differ by breast cancer case-control status and was close to 100%, but controls recalled use of statins slightly better than cases during that period (sensitivity for cases and controls: 83% and 93%, respectively).36 If these findings apply to the present study, such differential misclassification would have falsely decreased our risk estimates. However, it may not have an appreciable impact because the prevalence of statin use was relatively low in our sample and thus a small absolute number of cases and controls would have been misclassified as nonusers of statins.
Simvastatin, pravastatin, and fluvastatin were introduced in the US after 1991. Therefore, the present study could not evaluate long-term use of these agents, which may be etiologically relevant for cancer. Further, our results may be confounded if serum cholesterol levels are associated with breast cancer risk. However, recent results from a large prospective cohort study showed no association between serum cholesterol levels and breast cancer risk.21 We repeated our analysis among subjects that reported ever using any type of cholesterol-lowering drug and found that the risk estimates for statin users compared with users of other cholesterol-lowering drugs were decreased compared with the results from the analysis of the entire sample. However, the reference group in the restricted analysis was small (64 cases and 53 controls).
A strength of this study is the high proportion of eligible cases (80%) and controls (76%) that were successfully interviewed. However, our results may be biased if the prevalence of statin use among the controls was not the same as the populations from which the cases arose. Depending on whether controls that were successfully interviewed were more or less likely to use statins than eligible controls that were not successfully interviewed, our risk estimates may over- or underestimate the magnitude of the breast cancer risk associated with the use of statins.
In conclusion, in this study ever use of statins overall was not associated with breast cancer risk. Furthermore, there were no associations with lipophilic statins as a group or with the hydrophilic statin pravastatin. One significant inverse association, observed for fluvastatin, a lipophilic statin, may have emerged by chance. However, in view of the widespread use of these medications, further studies that consider associations with individual statins may be warranted.