The lipid-lowering agents known as statins are the top therapeutic class of prescription drugs in the United States. In 2005, Lipitor® (atorvastatin; Pfizer, New York, New York) was the number one drug both in the number of prescriptions dispensed (almost 80 million) and revenue sales (more than $7 billion).1 Data from the National Health and Nutrition Examination Surveys (NHANES) showed the prevalence of statin use tripled between surveys conducted in 1988–1994 and 1999–2002, with approximately 10% of all adults and nearly one-quarter of adults over 60 reporting statin use.2 The widespread use of statins reflect their efficacy in reducing levels of total cholesterol, low-density lipoprotein cholesterol, and triglycerides and increasing high-density lipoprotein cholesterol; their increased tolerability as compared with other antihyperlipidemic agents; and the high prevalence of hyperlipidemia as a result of the epidemic of overweight and obesity in this country.
Statins have a well-established role in reducing the incidence of adverse cardiovascular outcomes, including death, myocardial infarction, atrial fibrillation, and renal dysfunction.3–6 Studies of the mechanisms of statins and observations that cardiovascular benefits are experienced even in statin users without elevated cholesterol levels led to the recognition that actions of statins extend beyond their cholesterol-lowering properties.7,8 These “pleiotropic” effects of statins are not limited to the cardiovascular system. Indeed, their effects on cell proliferation have led to the hypothesis that statins may also reduce the risk of cancer. In vitro and in vivo studies have shown statins have multiple influences on tumor growth, including inhibition of cell growth by cell cycle arrest, induction of apoptosis, suppression of angiogenesis, and other mechanisms.9 The idea of a widely used class of drugs such as the statins having “two-for-the money” effects in reducing death from both cardiovascular disease and cancer certainly has great appeal. Unfortunately, the epidemiologic data accumulated to date have not yet provided convincing evidence to suggest that statins have a substantial effect on cancer incidence.
The relationship between statin use and cancer incidence has been evaluated in numerous observational studies and randomized clinical trials. Although there have been no clinical trials evaluating cancer as a primary outcome, cancer risk has been assessed in secondary analyses of many randomized controlled trials to assess the efficacy of statins for reducing cardiovascular morbidity and mortality. Several recently published meta-analyses have combined data from the nearly 3 dozen randomized controlled clinical trials of statins to evaluate cancer risk overall10–12 or for specific cancer sites.13,14 Relative risks (RRs) for overall cancer incidence in individual studies ranged from approximately 0.5 to 1.8, with the most extreme values coming from trials involving fewer than 500 patients. When the combined risk for all cancers was examined, no significant association was evident in either of the 2 most recent reviews, with one reporting a RR of 0.99 (95% confidence interval = 0.94–1.04)11 and the other a RR of 1.02 (0.97–1.07).10
Although randomized controlled clinical trials are generally considered the gold standard for establishing causality, the relationship between statins and cancer cannot be considered a closed issue. Despite the wealth of evidence from clinical trials, these studies had shortcomings that prevent one from making a definitive conclusion about whether statins neither increase nor decrease the risk for cancer. The randomized controlled trials were powered to detect cardiovascular outcomes, which are much more common than cancer outcomes and require shorter follow-up to accrue adequate number of events. Among the 35 randomized controlled trials included in the meta-analysis by Bonovas et al,11 only 5 studies had average follow-up times of 5 years or longer, and only one had 10 years of follow-up. Not surprisingly, with the limited follow-up time, the number of cancers in any individual study tended to be fairly small. Two-thirds of the studies had fewer than 100 incident cancer cases and only one study had more than 1000. Even when combining data from multiple studies, the ability to assess the risk of individual cancer sites was limited. Another concern about making conclusions from randomized controlled trials is the possibility that the stringent inclusion criteria for these trials may limit the generalizability of the findings to a more typical, postmarketing population of statin users.
Given the limitations of the randomized controlled trials for assessing the possible role of statins in cancer etiology, data from observational studies have the potential for furthering our understanding of this relationship. Some of the largest observational studies have shown reductions in overall cancer incidence ranging from 14% to 36%,15–18 with even larger reductions in risk for specific cancer sites (eg, a 47% reduction in colorectal cancer incidence).19 However, other studies showed either no significant association between statin use and cancer risk or an increase in risk.15,20,21
The study reported by Coogan and colleagues22 in this issue of Epidemiology is an important contribution to this literature. In their hospital-based, case–control study, 4913 cases and 3900 controls recruited between 1991 and 2005 were interviewed about lifetime medication use. Major advantages of this approach include the large number of cancer cases in their study and the inclusion of individuals with long-term exposure to statins. With nearly 5000 cancer cases, the investigators were able to evaluate overall cancer risk as well as the risk for 10 individual cancer sites. The lifetime medication history allowed the investigators to consider the risk of cancer by duration of statin use. Statins were first introduced to the U.S. market in the late 1980s. Therefore, with recruitment of cases and controls through 2005, there was the potential for including individuals who had exposures for well over 10 years. As shown in the paper, the prevalence of statin use showed a steady increase in the time period during which the cases and controls were enrolled. However, even with the size of this study, the number of long-term users was fairly small, limiting the conclusions one can make about the effects of statins when used for longer durations. Given current prescribing trends and the advocacy by some experts to target ever lower cholesterol levels, it is predictable that the number of long-term users in the population will only increase.
Certainly observational studies, whether case–control studies such as the one reported by Coogan and colleagues22 or prospective or retrospective cohorts, have many limitations. The probability of exposure misclassification because of the reliance on self-report by participants is a very real concern. Although it is sometimes recommended that pharmacoepidemiologic studies base exposure information on medical or pharmacy records, these methods also are prone to misclassification because individuals may not take all medicine that is prescribed or may obtain medications from multiple providers, making it logistically very challenging to capture all instances of medication use. Self-report of lifetime medication history, although imperfect, is a reasonable approach for drugs such as statins which are used on a chronic basis and are likely to be remembered accurately. Although the potential for errors in recall are lessened in cohort studies, a recent study by Jacobs et al20 pointed out that there have been important changes in statin use over time, highlighting the necessity of ascertaining exposure status at multiple points during follow-up.
The greater potential for confounding by lifestyle or medical characteristics in observational studies also has to be considered. Factors such as low physical activity, obesity, and high-fat/low-fruit-and-vegetable diets tend to cluster together, and may be related to both cancer risk and statin use. This suggests that careful assessment and analysis of confounders is critical for observational studies. Consideration of confounding by screening history is also important when assessing the association between statin use and cancer. By definition, a statin user is integrated into the medical care system. It is reasonable to assume that a patient being treated for cardiovascular disease prevention is more likely to receive preventive and screening care for other conditions. For example, a man who has been prescribed statins also may be more likely to receive prostate-specific antigen testing and be more likely to be diagnosed with prostate cancer. The failure to consider prostate cancer screening history could result in an apparent excess of prostate cancer cases among statin users and mask a protective effect of these drugs.
Given the current state of knowledge, what should be done regarding use of statins for cancer prevention? Clearly the current epidemiologic data do not warrant recommendations for statin use as a chemopreventive. Furthermore, the possibility of reducing cancer risk has to be balanced against the known adverse effects of these drugs. Although statins have a generally good safety profile, serious adverse events such as myopathies, renal dysfunction and hepatic dysfunction occur in up to 5% of patients.23 These adverse effects have to be given greater weight if there is any consideration of giving statins to generally healthy individuals for chemoprevention. Based on studies published to date, the risk/benefit ratio for statin use for cancer chemoprevention is not clear.
Although current data would not support recommending statin use for chemoprevention, the issue cannot be considered settled. Given the biologic plausibility of an anticancer effect of statins coupled with the limitations of published studies new studies involving adequate numbers of participants with long-term use of statins would be highly desirable. Could a randomized controlled trial be justified that is specifically designed to assess statins as chemopreventive agents, as has been suggested by some investigators?24,25 Such a randomized clinical trial would be an extremely expensive undertaking involving long-term follow-up of tens of thousands of volunteers. Based on the current conflicting evidence, it is hard to make the argument that such an undertaking is scientifically and fiscally justified. However, continued evaluation of the relationship between statins and cancer risk in observational studies seems prudent. In particular, further investigation is needed to determine whether effects on cancer incidence are related to dose, duration of use, extent of cholesterol lowering, or type of statin (hydrophilic versus lipophilic).
The relationship among cancer, lipid levels, and lipid-lowering drugs is another example of a situation in which the state of knowledge and the conventional wisdom have changed over time. When the first statin was introduced to the U.S. market in 1987, there were concerns that low total cholesterol was associated with increased risk for cancer.26,27 With accumulating evidence on the cellular effects of statins, the hypothesis emerged that statins could reduce cancer incidence. At present, there is no definitive answer on whether lipid-lowering drugs affect cancer risk. What is clear, however, is that any class of drugs that is being used by more than 10% of the adult population and one-quarter of the population over age 60, demands continued monitoring of its benefits and risks.
ABOUT THE AUTHORS
PATRICIA MOORMAN is an Associate Professor in the Department of Community and Family Medicine and ROBERT HAMILTON is an Associate in Research in the Department of Surgery (Urologic Surgery) at Duke University Medical Center. Dr. Moorman is an epidemiologist with longstanding interests in the etiologic relationships between medications and cancer. Dr. Hamilton is a urologist with a particular interest in the effects of statins on prostate cancer risk.
2. Carroll MD, Lacher DA, Sorlie PD, et al. Trends in serum lipids and lipoproteins of adults, 1960–2002. JAMA
3. Baigent C, Keech A, Kearney PM, et al, for the Cholesterol Treatment Trialists (CTT) collaborators. Efficacy and safety of cholesterol-lowering treatment: prospective meta-analysis of data from 90,056 participants in 14 randomised trials of statins. Lancet
4. Shepherd J, Cobbe SM, Ford I, et al, for the West of Scotland Coronary Prevention Study Group. Prevention of coronary heart disease with pravastatin in men with hypercholesterolemia. N Engl J Med
5. Collins R, Armitage J, Parish S, et al, for the Heart Protection Study Collaborative Group. MRC/BHF Heart Protection Study of cholesterol-lowering with simvastatin in 5963 people with diabetes: a randomized placebo-controlled trial. Lancet
6. Chan AW, Bhatt DL, Chew DP, et al. Early and sustained survival benefit associated with statin therapy at the time of percutaneous coronary intervention. Circulation
7. Bonetti PO, Lerman LO, Napoli C, et al. Statin effects beyond lipid lowering—are they clinically relevant? Eur Heart J
8. LaRosa JC. Pleitropic effects of statins and their clinical significance. Am J Cardiol
9. Chan KK, Oza AM, Siu LL. The statins as anticancer agents. Clin Cancer Res
10. Dale KM, Coleman CI, Henyan NN, et al. Statins and cancer risk: a meta-analysis. JAMA
11. Bonovas S, Filioussi K, Tsavaris N, et al. Statins and cancer risk: a literature-based meta-analysis and meta-regression analysis of 35 randomized controlled trials. J Clin Oncol
12. Bjerre LM, LeLorier J. Do statins cause cancer? A meta-analysis of large randomized clinical trials. Am J Med
13. Bonovas S, Filioussi K, Tsavaris N, et al. Use of statins and breast cancer: a meta-analysis of seven randomized trials and nine observational studies. J Clin Oncol
14. Dellavalle RP, Drake A, Graber M, et al. Statins and fibrates for preventing melanoma. Cochrane Database Syst Rev
15. Friis S, Olsen J. Statin use and cancer risk: an epidemiologic review. Cancer Invest
16. Blais L, Desgagne A, LeLorier J. 3-Hydroxy-3-methylglutaryl coenzyme A reductase inhibitors and the risk of cancer: a nested case-control study. Arch Intern Med
17. Graaf MR, Beiderbeck AB, Egberts AC, et al. The risk of cancer in users of statins. J Clin Oncol
18. Friis S, Poulsen AH, Johnsen SP, et al. Cancer risk among statin users: a population-based cohort study. Int J Cancer
19. Poynter JN, Gruber SB, Higgins PDR, et al. Statins and the risk of colorectal cancer. N Engl J Med
20. Jacobs EJ, Rodriguez C, Brady KA, et al. Cholesterol-lowering drugs and colorectal cancer incidence in a large United States cohort. J Natl Cancer Inst
21. Cauley JA, McTiernan A, Rodabaugh RJ, et al. Statin use and breast cancer: prospective results from the Women’s Health Initiative. J Natl Cancer Inst
22. Coogan PF, Rosenberg L, Strom BL. Statin use and the risk of ten cancers. Epidemiology
. 2006;in press.
23. Drug Facts and Comparisons
. 58th ed. St. Louis, MO: Wolters Kluwer Health; 2003;652–656.
24. Kumar AS, Benz CC, Esserman LJ. Clinical trials are required to prove the chemopreventive worth of statins. Arch Intern Med
25. Sprague J, Wood M. Breast cancer prevention: time for randomized controlled trials with statins. Arch Intern Med
26. Sherwin RW, Wentworth DN, Cutler JA, et al. Serum cholesterol levels and cancer mortality in 361,662 men screened for the Multiple Risk Factor Intervention Trial. JAMA
27. Knekt P, Reunanen A, Aromaa A, et al. Serum cholesterol and risk of cancer in a cohort of 39,000 men and women. J Clin Epidemiol