|Home | About | Journals | Submit | Contact Us | Français|
To determine whether 3-hydroxy-3-methylglutaryl coenzyme A inhibitors (statins) are associated with a decreased risk of colorectal cancer.
The population included 159,219 postmenopausal women enrolled in the Women’s Health Initiative in which 2000 pathologically confirmed cases of colorectal cancer were identified during an average of 10.7 (S.D. 2.9) years. Information on statins was collected at baseline and years 1, 3, 6, and 9. Self- and interviewer-administered questionnaires were used to collect information on other risk factors. Hazard ratios (HRs) and 95% confidence intervals (CIs) were calculated by the use of Cox proportional hazards regression to evaluate the relationship between statin use and risk. Statistical tests were two-sided.
Statins were used by 12,030 (7.6%) women at baseline. The annualized colorectal cancer rate was 0.13% among users and 0.12% among nonusers. The multivariable adjusted HR for users versus nonusers was 0.99 (95% confidence interval [CI], 0.83–1.20, p=.95), and 0.79 (95% CI, 0.56–1.11) for users of ≥3 years. In the multivariable adjusted time-dependent model, the HR for lovastatin was 0.62 (95% CI, 0.39–0.99). There was no effect of tumor location, stage or grade.
There was a reduction in colorectal cancer risk associated with lovastatin and a non-significant association with longer duration of use.
Colorectal cancer is the third-leading cause of cancer incidence and death among women in the United States, with an estimated 70,480 new cases and 24,790 deaths reported in 2010 (1). With the appropriate use of treatment, colorectal cancer is potentially curable, especially if discovered at an early stage; however, the clinical presentation can be insidious, and patients frequently present at an advanced stage (2). Although regular screening of asymptomatic patients by fecal occult blood testing, sigmoidoscopy, or colonoscopy is effective in decreasing the mortality associated with colorectal cancer, the majority of the population receives no screening or inadequate screening (3, 4), and rates of screening are much lower than for other common malignancies. The high death rate from colorectal cancer and inadequate acceptance of screening support the need to focus on chemoprevention as a way to more effectively impact colorectal cancer mortality.
A number of measures, including regular intake of aspirin or nonsteroidal anti-inflammatory drugs (5), calcium and vitamin D intake (6), and dietary changes including increased fiber (7) and whole grain intake (8), have been associated with lower colorectal cancer risk in observational studies; however, validation by randomized controlled trials have been disappointing (9, 10). Colorectal cancer involves the progression through well-defined morphological, cellular, and genetic events in the adenoma to carcinoma pathway (11). Statins are a logical candidate for chemoprevention in that they have pleiotropic effects in addition to cholesterol-lowering including, inhibition of rho GTPases (12–14), induction of apoptosis (12, 15–18), decrease in markers of chronic inflammation (19), inhibition of cell proliferation (20–22), and decrease in formation and progression of aberrant crypt foci (23). In a population based case-control study, Poynter et al. (24) demonstrated a 47% reduction in colorectal cancer risk among users of statins for 5 or more years. The relationship between statins and reduction in colorectal cancer risk has been reviewed in four recently completed meta-analyses (25–28). Overall, the authors of observational studies have shown a modest reduction in risk; however, these findings have not been demonstrated in randomized controlled trials.
The purpose of this study was to test whether statins and other lipid-lowering agents are associated with a lower risk of colorectal cancer among participants in the Women’s Health Initiative (WHI), which is the largest multicenter longitudinal study of postmenopausal women in the United States. In the WHI, detailed information on statin use and duration of use was collected at study entry, and additional follow-up information was collected at years 1, 3, 6, and 9. Cancer outcome data is available for an average (SD) of 10.7 (2.9) years of follow-up.
The WHI includes an observational study (OS; n = 93,676) and three clinical trials (CTs; n = 68,132) of hormone therapy, dietary modification, and calcium/vitamin D supplementation in postmenopausal women of mixed race and ethnicity (29). Recruitment was conducted between October 1, 1993, and December 31, 1998, at 40 clinical centers in the United States. Eligibility criteria included women ages 50–79 years who were postmenopausal, planned to remain in area where they were recruited, and had an estimated survival of at least 3 years. Study methods have been described in detail elsewhere (30, 31). Participants were followed through March 2005 and were invited to enroll in an extension study which lasted from April 2005 through September 2010.
The current analysis includes 91,912 women enrolled in the OS and 67,307 enrolled in the CT (159,219 total), and excludes 963 women with a previous diagnosis of colorectal cancer, 1624 women with an unknown previous history of colorectal cancer, and 2 women with unknown information on previous statin use. All participants signed informed consent forms, and all protocols and procedures were approved by institutional review boards of the participating institutions. Follow-up was through August 14, 2009, for a mean (SD) follow-up of 10.7 (2.9) years and a maximum of 15.6 years.
At study entry, participants brought in all of their current prescriptions and each medication name was directly entered by study personnel into the WHI database, which assigned national drug codes by the use of Medispan software (First DataBank, Inc., San Bruno, CA). Participants also reported duration of use for each medication. Information on prescription medications was similarly updated at years 1, 3, 6, and 9 in the CT, and at year 3 in the OS.
Statins were defined as any 3-hydroxy-3-methylglutaryl coenzyme A reductase inhibitors and were classified on the basis of solubility in octanol (lipophilicity) or water (hydrophilicity) (32, 33). Lipophilic (or hydrophobic) statins (lovastatin, simvastatin, fluvastatin, cerivastatin) penetrate the plasma membrane whereas hydrophilic statins (pravastatin, atorvastatin and rosuvastatin) do not (34–36). Statins were classified by potency on the basis of lipid-lowering efficacy as low (fluvastatin and lovastatin), medium (pravastatin), and high (simavastatin, atorvastatin, cerivastatin and rosuvastatin) (34, 35, 37). Data were collected on other lipid-lowering medications, including fibrates, colestipol, probucol, cholestyramine, niacin, and nicotinic acid.
Cancer diagnoses were updated annually in the OS or semiannually in the CT by mail and/or telephone questionnaires. Participant or next-of-kin reports of colorectal cancer were verified by centrally trained physician adjudicators after review of medical records and pathology reports using the Surveillance Epidemiology and End Results coding system (38). Only invasive colorectal cancer cases were included and the following unusual or rare histologic types were censored: adenocarcinoma in the setting of polyposis coli (n = 1), malignant carcinoid tumor (n = 14), neuroendocrine carcinoma (n = 9), infiltrating ductal carcinoma, NOS (n = 2), medullary carcinoma (n = 1), and malignant melanoma, not otherwise specified (n = 3). Information on the frequency of screening tests, including fecal occult blood tests, rectal examinations, and sigmoidoscopy or colonoscopy was collected at baseline and updated semiannually in the CT and annually in the OS. Screening rates were not protocol defined.
Baseline questionnaires were used to collect information on race or ethnicity (White, African American, Hispanic, Native American, Asian/Pacific Islander, or unknown), physician-diagnosed diabetes, high serum cholesterol that required treatment with pills, history of coronary artery disease, educational level (<high school, high school diploma/GED, or O high school diploma/GED), family colorectal cancer history, use of nonsteroidal anti-inflammatory drugs or aspirin (yes/no), current and past smoking, and physical activity in metabolic equivalents. Alcohol consumption (none/past drinker, <1 drink/week, or ≥1 drink/week), percentage of calories from fat (≥30% versus <30% of calories from fat) and other dietary data were estimated from the WHI food-frequency questionnaire (39). Body mass index (BMI) was calculated as weight in kilograms divided by height in meters squared.
Current and previous use of menopausal hormones and oral contraceptives were ascertained by interviewer administered questionnaires that queried type, route of administration, number of pills per day/week, and duration for each preparation ever taken. Hormone therapy users were defined as those who used estrogen (with or without progestin) for at least 3 months after menopause.
Characteristics of statin users at baseline were compared with those of nonusers by χ2 tests. Annualized rates of colorectal cancer were calculated as the percentage of women with an event divided by total follow-up time in years by statin use categories and other lipid-lowering agents at baseline. An a priori plan specified that we perform selected subgroup analyses by duration (<1 year, 1 to <3 years, and ≥3 years), type, potency, and hydrophobic status. Women who used two or more statins were included in analyses that compared statin use to none, but were excluded from analyses that examined details of use (by type, potency or lipophilic). Separate analyses were conducted for proximal, distal and rectal sites as well as other clinical characteristics including stage, tumor size, lymph node involvement and grade. Hazard ratios (HRs) for colorectal cancer among statin users versus nonusers, and 95% confidence intervals (CIs) were computed from Cox proportional hazards analyses. Tests for the proportional hazards assumptions were conducted by a Cox model that included statin use and the interaction of statin use with follow-up time, and testing for a zero coefficient on the interaction term. Results of these analyses showed that the assumptions were not violated.
Cox proportional hazards analyses were used to assess associations between statins and colorectal cancer. Two models were developed including an age-adjusted model and a multivariable-adjusted model. Both sets of models were stratified as to allow the baseline hazard function to vary by assignment to active hormone or placebo in the two WHI hormone trials (estrogen plus progestin and estrogen alone), assignment to intervention or control in the dietary modification trial, or enrollment in the OS, and/or WHI extension study participation. To control for confounding, the multivariable model was additionally adjusted for age, ethnicity, education, smoking, alcohol use, physical activity, BMI, percent energy from fat, fruit and vegetable intake, dietary calcium, calcium supplement use, selenium supplement use, current healthcare provider, last medical visit within one year, colon screening at baseline, current HT use, family history of colorectal cancer, history of colon polyp removal, use of nonsteroidal anti-inflammatory drugs, hypertension, history of stroke and history of coronary artery disease.
To evaluate the effect of the change in statin use over time, we conducted time-dependent models in which we incorporated updated information on statins (whether or not participants had started statins, statin type, category and potency) at years 1, 3, 6, and 9 in the CT and year 3 in the OS. We conducted time-dependent Cox models to examine the effect of colon screening during the study, where any report of screening (rectal examination, hemoccult guaiac, colonoscopy, sigmoidoscopy, flexible sigmoidoscopy, or barium enema x-ray), was incorporated on a yearly basis. Comparisons of colorectal cancer tumor characteristics between statin users and nonusers were based on χ2 and Fisher exact tests. We evaluated interaction effects of age, family history, BMI, colonoscopy screening, and hormone therapy with statin use on the association with colorectal cancer. All analyses were conducted with Statistical Analysis Systems (SAS) software, version 9.2 (SAS Institute, Inc., Cary, NC). All statistical tests were two-sided with a significance level of .05.
The WHI cohort consisted of 12,030 statin users (7.6%)at baseline. Table 1 shows characteristics of WHI participants by statin use. Although most of the absolute differences between statin users and nonusers were small, many were statistically significant because of the large sample size. Statin users were more likely to be older than nonusers (mean age, SD, 65.6 [6.5] and 63.0 [7.2] years, respectively) and to have a greater BMI (28.9 [5.5] and 27.9 [6.0] kg/m2). Statin use was associated with tobacco consumption, previous colon screening, a family history of colorectal cancer, one or more co-morbid medical conditions, and use of nonstatin lipid-lowering medications or aspirin. Non-use of statins was associated with greater levels of education and family income, greater alcohol intake, more physical activity, use of hormone therapy and a diet with >30% calories from fat.
Table 2 shows the distribution of statin users by type of statin, potency, lipophilicity, and duration of use. Simvastatin was the most commonly used single statin, and the majority of women used either a low- or a high-potency statin and/or a statin classified as lipophlilic. Among statin users at baseline, 3980 (33.1%) took statins for less than 1 year, 4088 (34) took statins for 1–3 years, and 3962 (32.9%) took statins for 3 or more years. Table 3 shows the incidence of invasive colorectal cancer (annualized %) and HRs by statin use and other lipid-lowering medications. There were 2000 women diagnosed with invasive colorectal cancer with a yearly incidence of 0.13% for statin users compared to 0.12% for non-users. There were no significant differences in risk of colorectal cancer for statin versus nonstain users at baseline in the age and WHI trial adjusted model (HR, 0.95; 95% CI, 0.81–1.12) or in the multivariable adjusted model (HR, 0.99; 95% CI, 0.83–1.20).
There were no significant differences in risk on the basis of type of statin, potency or category. There was a 21% decrease in colorectal cancer risk for statin use of >3 years’ duration (HR, 0.79; 95% CI, 0.56–1.11) but this difference was not statistically significant (p = .13). There was no significant association for other lipid-lowering medications and colorectal cancer risk. There was also no significant association between risk of proximal, distal, or rectal cancer and statin use, and no significant association between statins and colorectal cancer risk for any of the observed tumor characteristics at diagnosis (Table 4). There were no other significant interaction effects (data not shown).
When statin use reported at years 1, 3, 6, and 9 was incorporated into a multivariable time-dependent model, there was no overall effect on colorectal cancer risk (HR, 1.02; 95% CI, 0.89–1.16; Table 5). There was no significant association with colorectal cancer risk by statin potency or category or with other lipid-lowering medications in the time dependent models. However, when statin type was taken into account, there was a marginally significant lower risk associated with use of lovastatin in the multivariable model (HR, 0.62; 95% CI, 0.39–0.99, p = .05).
We hypothesized that statins are associated with a lower risk of colorectal cancer on the basis of in vitro and in vivo data suggesting that the mechanism of anticancer effects are through inhibition of small GTPases (Ras and Rho) (12–14) induction of apoptosis (12, 16–18, 25), and regression of aberrant crypt foci (23). The authors of previous epidemiologic studies have reported an association of statins with either a reduction in risk of cancer overall (40, 41), or specifically a reduction in colorectal cancer risk (27, 28). In our analysis, we found no overall protective effect of statins, or when statins were considered by potency or category; however, we observed a significant reduction in colorectal cancer risk for lovastatin specifically in a time-dependent analysis, and a modest, although not significant reduction for overall statin use of ≥3 years. We also found no association of colorectal cancer risk with use of nonstatin lipid-lowering agents, although only 1.4% of the cohort reported use of these medications at baseline.
Results from previous studies have been mixed. In a 2007 meta-analysis of 18 studies involving more than 1.5 million patients, there was no significant association between statins and colorectal cancer in six randomized controlled trials, and in 3 cohort studies, although among the nine case-control studies cited (including two studies presented as abstracts), there was an overall modest reduction in risk (26). This risk reduction was mainly attributed to the findings of a population-based study completed by Poynter et al. (24) in which statin use was compared among 1953 cases and 2015 controls in Northern Israel. Their results showed that statin use of 5 or more years was associated with a 47% reduction in risk.
In a recently updated meta-analysis of 11 randomized controlled trials, 13 case-control, and 8 cohort studies, authors estimated the overall effect of statins and reported an 8% reduction in risk, although again there were no significant effects seen in randomized controlled trials (25). It should be noted that the randomized trials were designed to assess the impact of statins on cardiovascular health and were not powered to assess the role of statins in cancer prevention. In a review of eight cohort studies, the authors revealed heterogeneous findings with one study showing a significantly increased risk (42), two other studies showing a reduction in risk (43, 44), and five others showing no association (45–49). Overall cohort studies revealed a non-significant marginal reduction in risk among users of statins (HR. 0.89; 95% CI, 0.75–1.05) (25).
Our results showing a marginal reduction in risk of colorectal cancer associated with statins are consistent with other cohort studies (25). The reduction in risk associated with lovastatin in the time-dependent analyses though was of marginal significance, and could have been due to chance because of multiple comparisons. It is of note that at baseline, lovastatin was the second most commonly used statin in the WHI and accounted for 47% of individuals that used statins for more than 3 years. By study year 9, lovastatin accounted for only 5% of those who used statins for more than 3 years, 55% of whom reported taking atorvastatin (data not shown). It is also possible that the results by duration of use, did not reach statistical significance because of the relatively uncommon baseline use of statins in the WHI for women enrolling from 1993 through 1998 (7.6%), compared with other studies in which the prevalence of statin use ranges from 1.8% to 76%. Also, in the WHI there were only a few cases (n = 46) who used statins for 3 or more years, and additional analyses of statin use of 5 or more years reported in 26 women were not significant (data not shown). It is of note that the WHI analysis of statins and breast cancer risk revealed a reduction in breast cancer risk associated with lovastatin as well as simvastatin and fluvastatin which are also both lipophilic statins (50); however, simvastatin and fluvastatin were not associated with a reduction in colorectal cancer risk in the current analysis.
The strengths of the WHI include the prospective design; information on statin use through multiple years of followup; inclusion of a large, racially diverse sample of well-characterized women; large number of colorectal cancer cases; collection of detailed information on a comprehensive range of risk factors; complete follow-up of outcomes; assessment of screening at baseline and follow-up; blinded adjudication of colorectal cancer via pathology report review; description of colorectal cancer clinical characteristics; and the ability to examine associations by statin category. Limitations include the relatively low prevalence of statin use, lack of information on dose, and low power to examine long-term effects. Although statin use was determined by self-report, the actual data were derived from medications that WHI participants brought in to their clinic visits and were directly recorded by study personnel. Other limitations include the fact that there may be residual confounding by unmeasured factors and that participants were not required to have colorectal cancer screening.
In conclusion, despite sound scientific plausibility, the association of statins with a reduced risk of colorectal cancer is not clearly evidenced in the WHI cohort. Although the reduction in risk associated with lovastatin in the time-dependent analysis is provocative, and the marginal reduction in risk with longer term exposure was consistent with the literature, our results are still not conclusive. Recent studies suggest that the efficacy of statins in reducing both cardiovascular (51) and colorectal cancer risk (52) may be related to genetic variation in 3-hydroxy-3-methylglutaryl coenzyme A reductase activity. The results from these studies support the need for randomized trials of statin use among individuals at high risk for colorectal cancer based on family history or genetic predisposition.
The WHI program is funded by the National Heart, Lung, and Blood Institute, National Institutes of Health, U.S. Department of Health and Human Services through contracts N01WH22110, 24152, 32100-2, 32105-6, 32108-9, 32111-13, 32115, 32118-32119, 32122, 42107-26, 42129-32, and 44221. The study was also funded by the Cancer Center Support Grant NIH: NCI P30CA022453.
National Heart, Lung, and Blood Institute, Bethesda, MD: Jacques Rossouw, Shari Ludlam, Joan McGowan, Leslie Ford, and Nancy Geller.
Clinical Coordinating Center
Fred Hutchinson Cancer Research Center, Seattle, WA: Ross Prentice, Garnet Anderson, Andrea LaCroix, Charles L. Kooperberg; Medical Research Labs, Highland Heights, KY: Evan Stein; University of California at San Francisco, San Francisco, CA: Steven Cummings.
Albert Einstein College of Medicine, Bronx, NY: Sylvia Wassertheil-Smoller; Baylor College of Medicine, Houston, TX: Haleh Sangi-Haghpeykar; Brigham and Women’s Hospital, Harvard Medical School, Boston, MA: JoAnn E. Manson; Brown University, Providence, RI: Charles B. Eaton; Emory University, Atlanta, GA: Lawrence S. Phillips; Fred Hutchinson Cancer Research Center, Seattle, WA: Shirley Beresford; George Washington University Medical Center, Washington, DC: Lisa Martin; Los Angeles Biomedical Research Institute at Harbor–UCLA Medical Center, Torrance, CA: Rowan Chlebowski; Kaiser Permanente Center for Health Research, Portland, OR: Erin LeBlanc; Kaiser Permanente Division of Research, Oakland, CA: Bette Caan; Medical College of Wisconsin, Milwaukee, WI: Jane Morley Kotchen; MedStar Research Institute/Howard University, Washington, DC: Barbara V. Howard; Northwestern University, Chicago/Evanston, IL: Linda Van Horn; Rush Medical Center, Chicago, IL): Henry Black; Stanford Prevention Research Center, Stanford, CA: Marcia L. Stefanick; State University of New York at Stony Brook, Stony Brook, NY: Dorothy Lane; The Ohio State University, Columbus, OH: Rebecca Jackson; University of Alabama at Birmingham, Birmingham, AL: Cora E. Lewis; University of Arizona, Tucson/Phoenix, AZ: Cynthia A. Thomson; University at Buffalo, Buffalo, NY: Jean Wactawski-Wende; University of California at Davis, Sacramento, CA: John Robbins; University of California at Irvine, CA: F. Allan Hubbell; University of California at Los Angeles, Los Angeles, CA: Lauren Nathan; University of California at San Diego, LaJolla/Chula Vista, CA: Robert D. Langer; University of Cincinnati, Cincinnati, OH: Margery Gass; University of Florida, Gainesville/Jacksonville, FL: Marian Limacher; University of Hawaii, Honolulu, HI: J. David Curb; University of Iowa, Iowa City/Davenport, IA: Robert Wallace; University of Massachusetts/Fallon Clinic, Worcester, MA: Judith Ockene; University of Medicine and Dentistry of New Jersey, Newark, NJ: Norman Lasser; University of Miami, Miami, FL: Mary Jo O’Sullivan; University of Minnesota, Minneapolis, MN: Karen Margolis; University of Nevada, Reno, NV: Robert Brunner; University of North Carolina, Chapel Hill, NC: Gerardo Heiss; University of Pittsburgh, Pittsburgh, PA: Lewis Kuller; University of Tennessee Health Science Center, Memphis, TN: Karen C. Johnson; University of Texas Health Science Center, San Antonio, TX: Robert Brzyski; University of Wisconsin, Madison, WI: Gloria E. Sarto; Wake Forest University School of Medicine, Winston-Salem, NC: Mara Vitolins; Wayne State University School of Medicine/Karmanos Cancer Institute, Detroit, MI: Michael S. Simon.
Women’s Health Initiative Memory Study
Wake Forest University School of Medicine, Winston-Salem, NC: Sally Shumaker.
This study was presented in part at the 2010 AACR Frontiers in Cancer Prevention, Philadelphia, PA.