Reference
McDougall JA, Malone KE, Daling JR, Cushing-Haugen KL, Porter PL, Li CI. Long-term statin use and risk of ductal and lobular breast cancer among women 55 to 74 years of age. Cancer Epidemiol Biomarkers Prev. 2013;22(9):1529-1537.
Design
The authors used data from a population-based case-control study of breast cancer conducted in the Seattle-Puget Sound region to investigate the relationship between long-term statin use and breast cancer risk. Women diagnosed with invasive ductal carcinoma (IDC) or invasive lobular carcinoma (ILC) were compared with control women. All participants were interviewed in person, and data on hypercholesterolemia and all episodes of lipid-lowering medication use were collected through a structured questionnaire. The relationship between statin use and IDC and ILC risk was assessed using polytomous logistic regression.
Participants
In total, 916 invasive ductal carcinoma (IDC) and 1,068 invasive lobular carcinoma (ILC) cases in patients 55 to 74 years of age diagnosed between 2000 and 2008 were compared with 902 control women.
Key Findings
Women currently using statins and who had been taking them for 10 years or longer had an 83% higher risk of getting IDC (95% confidence interval [CI]: 1.14–2.93) and nearly double the risk of ILC (95% CI: 1.25–3.12) compared with women who never used statins. These associations were stronger when limited to cases and controls with hyperlipidemia, with a doubled risk of IDC (95% CI 1.2–3.6) and a 2.4-fold risk of ILC breast cancer (95% CI: 1.4–4.2) compared to participants who never used statins.
Practice Implications
About 25% of the women in the United States over age 45 are currently taking statin drugs to prevent stroke and heart disease. These drugs are effective at lowering high levels of blood lipids and by doing so, lowering risk of cardiovascular disease.
A number of papers published over the last 2 years have suggested that statins may also reduce the risk of patients dying from cancer, in particular breast cancer. If so, it would be good that so many women are already taking these drugs. This current study by McDougall et al contradicts the earlier papers and suggests that the opposite may be true. If so, all these women taking statin drugs is not a good thing. Knowing whether statins will inhibit cancer or increase cancer is important given the widespread use of these drugs. Which is it?
This isn’t the first time our thinking on statins and cancer has taken a turn and reversed itself. When Alsheikh-Ali et al (2007) first reported a significant association between low levels of low-density lipoprotein (LDL) cholesterol and cancer among patients taking statins, the initial concern was that the statins were causing cancer. This was a meta-analysis of 23 drug trials, yielding 309,506 person-years of follow-up—not a small study. Alsheikh-Ali reported that “the risk of cancer is significantly associated with lower achieved LDL-C levels.”1
This wasn’t actually the case. A year later in a second study, Alsheikh-Ali amended the initial conclusion: “There is no evidence that statin use causes cancer, although patients who reduce their LDL cholesterol level with statins appear to have a significantly increased risk of the disease.”
Alsheikh-Ali’s first paper used only data from patients taking statins. The second paper included data from both people taking statins and from people not taking them. Fifteen randomized controlled trials provided data from 97,000 patients yielding 437,000 person-years of follow-up.
During these trials, 5,752 patients developed a new cancer. The incidence was 4%–27% per 1,000 person-years for those taking statins and 6%–24% per 1,000 person-years for the people not taking statins, the same risk for both groups. The people who had low LDL cholesterol levels had an increased risk for cancer. This was true even for the patients taking placebo and not statins. The statins were not to blame. Low LDL cholesterol levels may be just a warning that cancer is developing. As one researcher noted, "We know that cancers can significantly lower cholesterol levels as much as 10 years before they surface clinically."2 Cancer cells consume LDL cholesterol faster than normal cells. Perhaps statins were not the culprit, but rather the cancer itself was.
In our rush to help people, we have to be careful not to hurt them. Risk of doing a patient harm always trumps unproven benefit.
Then 2 years ago, the tide turned, and statins were suddenly good. The shift started with Nielsen’s 2012 study that reported people in Denmark who took statins had a lower risk of dying from cancer than non–statin users. The researchers had assessed causes of mortality among all Danes diagnosed with cancer between 1995 and 2007 and followed them through 2009. Of patients 40 years or older, 18,721 had used statins regularly before the cancer diagnosis and 277,204 had never used statins. Statin-users had a 15% lower risk of dying from any cause and from cancer. Reduced cancer deaths among statin-users were seen for 13 cancer types.3
Then in April 2013, Murtola reported that statin use was associated with a 66% reduction in the risk of dying from breast cancer.4 Murtola had conducted a retrospective study looking at statin use and breast cancer mortality among the 31,114 women with breast cancer who were diagnosed in Finland between 1995 and 2003. During follow-up, 6,011 of the women died, 3,169 due to breast cancer. The death rate among statin-users was 7.5% while among non–statin users it was 21%.
Women with localized disease who took statins were 67% less likely to die than nonusers (hazard ratio [HR], 0.33). Among those with metastatic disease, statin use was associated with a 48% decreased risk of death (HR: 0.52). Finland’s national health database allowed calculation of HR risk by the type of statin taken, including simvastatin (HR: 0.47), atorvastatin (HR: 0.27), fluvastatin (HR: 0.35), and pravastatin (HR: 0.50). Median follow-up was about 3 years but ranged from less than 1 year to 9 years.
These are the kind of numbers that get our attention. Suddenly statins, in particular atorvastatin, could be considered a powerful therapy in breast cancer treatment.
In the months since these two positive studies were published, a number of other studies have also shown benefit. Brewer et al (2013) reported that hydrophilic statins improved progression-free survival in patients with inflammatory breast cancer (IBC). They analyzed data from 723 IBC patients treated at the MD Anderson Cancer Center in Houston from 1995 to 2011. Statins were classified by Ahern’s system ranking them from hydrophilic to lipophilic (H-statins vs L-statins). H-statins use was associated with significantly improved progression-free survival compared with no statin (HR: 0.49; 95% CI: 0.28–0.84, P<.01).5
Other in vitro and animal experiment reported positive effects such as increasing breast cancer apoptosis,6 preventing carcinogenesis,7 inhibiting tumor growth,8 and inhibiting growth of triple negative breast cancer cells.9
Unfortunately not all trials report clear benefit.
A clinical trial that gave lovastatin (40 mg 2x/d/6 mo) to women at high risk of breast cancer found no significant change in breast duct cytology or other biomarkers of breast cancer risk.10 A study that compared 565 breast cancer cases with 2,260 controls found no significant differences in breast cancer risk between women who took statins and those who did not.11
A large German study was also inconclusive. Data from 3189 patients with invasive breast cancer from stage I to stage IV and 3024 patients with breast cancer from stage I to stage III were analyzed for recurrence risk. During a median 5.3 years of follow-up, 404 of 3189 stage I to stage IV patients died, and 286 deaths were attributed to breast cancer. While statistically nonsignificant, use of lipid-lowering drugs was associated with an increase in non–breast cancer mortality (HR: 1.49; 95% CI: 0.88–2.52) and increased overall mortality (HR: 1.21; 95% CI: 0.87–1.69). Limiting data to only stage I to stage III breast cancer patients, lipid-lowering drug use was nonsignificantly associated with a reduced risk of recurrence (HR: 0.83; 95% CI: 0.54–1.24) and of reduced breast cancer-specific mortality (HR: 0.89; 95% CI: 0.52–1.49).12 Nonsignificant results, even if trending in a good direction, are still nonsignificant and probably should be ignored.
A particularly large meta-analysis on statin use and the risk breast cancer risk found no significant effect. The researchers reported,
A total of 24 (13 cohort and 11 case-control) studies involving more than 2.4 million participants, including 76,759 breast cancer cases contributed to this analysis [;] Statin use and long-term statin use did not significantly affect breast cancer risk (relative risk [RR] = 0.99, 95 % CI = 0.94, 1.04 and RR = 1.03, 95 % CI = 0.96, 1.11, respectively).13
Another meta-analysis analyzed data drawn from 22 randomized controlled trials with 66,582 patients receiving statins and 66,604 receiving placebo. Five years of statin therapy had no effect on the risk of cancer-related death (RR: 1.00; 95% CI: 0.93, 1.08).14 Possibly the largest prospective study on this topic reported “no relationship between statins and breast cancer risk.” Using information from the Women's Health Initiative (WHI) yielded data from 154,587 postmenopausal women, with 7430 confirmed breast cancer cases. Of these women, 11,584 (7.5%) used statins at baseline. The research team concluded, “The annualized rate of breast cancer was 0.42% among statin users and 0.42% among nonusers.”15
If the effect was as large as Nielsen and Murtola had reported, why wasn’t it showing up in these other studies? According to Santa-Maria and Stearns in a 2013 analysis, when considered as a group, the research studies unfortunately "have not identified a strong relationship between statin use and reduced breast cancer incidence. These studies have several limitations and were not designed to detect modest effects in high-risk populations. Additional focused epidemiological and translational studies in high-risk populations are needed to justify and guide definitive large prospective trials."16
How could Nielsen’s data have been so far off? One explanation suggested is that in Nielsen’s study, women taking statins were more likely to have cardiovascular disease (70% vs 21%) and diabetes (18% vs 3%) than non–statin users. This could have led to disproportionately higher use of aspirin and or metformin in the statin-users. Both medications are associated with reduced cancer-related mortality. Nielsen did look again at the data with possible aspirin use in mind and eliminated all participants with cardiovascular disease (the only indication in Denmark for routine aspirin use), and this second analysis yielded the same results.17
Perhaps statins do better only in observational studies but fail in randomized controlled trials. Healthy-user bias in observational studies may explain this difference. Doctors may unconsciously but selectively underprescribe statins to obese patients or smokers because of their unhealthy lifestyles.18 Healthier patients might be the ones who end up taking statins.
We also have to ask another question: If statin use is associated with such a significant increased risk of breast cancer as seen in the McDougall study, why wasn’t it apparent in these other studies? The results are so inconsistent that we must question whether we are blind to some confounder. We may be grouping too many different genetically diverse types of cancer together under the blanket term breast cancer. Some of these cancers may be more sensitive to statins than others. For example, some tumors express HMG-CoA reductase (HMGCR), and these tumors may be the ones that respond to statins.
Swedish data from 2013 suggest that testing HMGCR can predict if cancers will respond to statin treatment. Tissue samples from 50 women diagnosed with invasive breast cancer and given high-dose atorvastatin (80 mg/d) for 2 weeks before cancer surgery were compared before and after statin therapy. Ki-67 expression between paired samples decreased nonsignificantly by only 7.6% after statin treatment (P=.39) but in tumors expressing HMGCR, the rate-limiting enzyme of the mevalonate pathway, Ki-67 dropped a significant 24% (P=.02). Statins have the most effect on HMGCR-positive tumors.19
There may be other ways to predict which tumors will respond to statins. A 2014 paper reports on a study in which 19 different breast cancer cell lines were treated with a statin drug (fluvastatin), and this yielded a range of responses; cell death was triggered only in a subset of sensitive cell lines, and this response was associated with an estrogen receptor alpha (ERα)-negative, basal-like tumor subtype. Using these data, researchers claim they are developing a gene signature test that will predict statin sensitivity.20
We may also be grouping too many different drugs together under the term statin. The Ahern system for classifying statins as H-statins or L-statins based on whether they are hydrophilic or lipophilic may help. The more lipophilic the statin, the greater impact it has on nonhepatic tissues. The more hydrophilic the statin is, the greater the impact on the liver. Thus H-statins are thought to reduce adverse drug effects.21 Simvastatin is among the most lipophilic of the statins while pravastatin is one of the least.
Using this classification system, Harvard researchers reported that after analyzing data on all the women in Denmark diagnosed with invasive breast cancer between 1996 and 2003 (18,769 in total), cancer rates varied by the type of statin the women took. Women who used simvastatin, a lipophilic statin, had approximately 10 fewer breast cancer recurrences per 100 women after 10 years of follow-up compared with women who were not prescribed a statin. Women who took H-statins had about the same risk of breast cancer recurrence as women not taking statins. Thus the type of statin could matter, as it is only the L-statins that act.22 This may in part explain the high frequency of breast cancer for women taking pravastatin in the Cholesterol and Recurrent Events (CARE) cohort as this drug is an H-statin. Women in this trial who received pravastatin were 12 times as likely to have breast cancer as controls receiving placebos.23 Perhaps if all of these studies paid more attention to the particular statin drugs the women were taking, the results might make more sense.
There is also the issue of 27-hydoxycholesterol (27HC) discussed in a recent issue of the Natural Medicine Journal. [link] Several recent studies link cholesterol metabolism to breast cancer. A specific cholesterol metabolites can promote breast cancer.24,25
This cholesterol metabolite, 27HC, is of interest because in several animal models, it interacts with the estrogen receptor and increases breast cancer growth and metastasis. Greater amounts of the enzyme that breaks down 27HC are associated with improved patient survival. Increased amounts of the enzyme that converts cholesterol to 27HC are found in more aggressive breast tumors. There is more 27HC in breast tumors than in healthy breast tissue of the same patients and far more 27HC in patients with cancer than in patients without. Some statin drugs may interfere with breast cancer because they prevent conversion of cholesterol to 27HC. 26 Others may not. At this point we do not know.
A March 2014 editorial sums up the current situation surrounding the research on statins and breast cancer:
As of today, the accumulated epidemiological evidence does not support the hypothesis that statin use affects the risk of developing breast cancer when taken at low doses for managing hypercholesterolemia. However, current evidence cannot exclude an increased risk of breast cancer with statin use in subsets of individuals, for example, the elderly. On the other hand, some studies show that statins might be useful to prevent recurrence and improve survival in patients already suffering from certain breast cancer types. They could also be combined with certain anticancer drugs and potentiate their effects, ameliorate their side effects or prevent the development of resistance. Further research is warranted to clarify these issues.27
While the results reported by both Nielsen and Murtola were exciting when published, McDougall et al should have a sobering influence. In our rush to help people, we have to be careful not to hurt them. Risk of doing a patient harm always trumps unproven benefit.
