Statins Cut Colon Ca Risk in Half in Retrospective Study

NEW ORLEANS—In a retrospective, observational study of nearly 4,000 patients in northern Israel, statin use for at least 5 years reduced the risk of colorectal cancer by 46%, after adjustment for known protective factors, including use of aspirin and nonsteroidal anti-inflammatory drugs (NSAIDs). The effects seen do not appear to be the result of a reduction in cholesterol levels. While the investigators, from the Molecular Epidemiology of Colorectal Cancer (MECC ) Study Group, cautioned that it is premature to change either the standard of care for colorectal cancer or the indications for statins, they said statins merit further investigation in this area.

Jenny N. Poynter, MPH, of the University of Michigan, presented the study results at the 40th Annual Meeting of the Society of Clinical Oncology (abstract 1), on behalf of the MECC Study Group. MECC is led by Stephen B. Gruber, MD, PhD, MPH, director of clinical cancer genetics, University of Michigan Comprehensive Cancer Center, and Gadi Rennert, MD, director of the National Cancer Control Center for Clalit Health Services, Israel’s largest HMO.

The MECC study, initiated in 1999 with a $4.8 million grant from the National Cancer Institute and additional funding from the Irving Weinstein and Ravitz Foundations, explores how genes, diet, exercise, and other factors interact to produce colon cancer.

Given that the descendants of Ashkenazi Jews have unusually high rates of colon cancer (with a novel cancer susceptibility allele, APC I1307K, identified in 6% and appearing to double colorectal cancer risk), the researchers believed an Israeli-based population study including this ethnic group, as well as Arab patients and Jews of non-Ashkenazi ethnicity, might provide interesting information about possible nongenetic factors influencing development of this disease.

Statins were assessed in this population, Ms. Poynter said, "because they inhibit HMG CoA reductase, which is overexpressed in colorectal cancer cell lines, and induce apoptosis in colorectal cancer cell lines at physiologically relevant concentrations. They have also been shown to reduce tumor formation in animal models of colorectal cancer."

Eligible patients were those diagnosed with colon or rectal cancer between 1998 and 2004 (n = 1,814 cases). Controls (n = 1,959) were matched to cases based on age, sex, and Jewish vs non-Jewish ethnicity. The researchers also performed matched analyses on 1,570 perfectly matched case-control pairs.

Medication use, including statins, was assessed in a structured in-person interview. Classified as "statin users" were those reporting use of any statin for at least 5 years; nonusers had no reported statin use. Regular statin use (prescriptions filled more than three times per year) was validated from prescription records from the Clalit Health System database for the majority of patients who reported using these drugs. All pathology was independently confirmed by a single pathologist at the University of Michigan.

Reduced Cancer Prevalence

Use of statins for at least 5 years was reported by 328 participants: 106 cases (5.8%) and 222 controls (11.3%). "In an unadjusted, unmatched analysis, we saw a strong protective effect of statins, with an odds ratio (OR) of 0.49, and a 95% confidence interval from 0.38 to 0.62, and a P value that was highly significant [P < .0001]," Ms. Poynter said. "Similar results were seen in a matched analysis of the 1,570 pairs currently available."

Pravastatin (Pravachol) and simvastatin (Zocor) were the two most common statins used, accounting for 44% and 52% of use, respectively. Both drugs were associated with a reduced risk of colorectal cancer (OR = 0.45 for prava-statin, 0.47 for simvastatin).

The only other cholesterol-lowering drug commonly used in this study was the fibric acid derivative bezafibrate (Bezalip), which has pharmacologic activity distinct from the statins and lowers cholesterol by increasing activity of lipoprotein lipase. In the study, 39 patients reported using bezafibrate for at least 5 years. No significant association was seen between bezafibrate use and incidence of colorectal cancer (OR = 1). "Thus, the protective effect of statins is not likely to be due to a general effect of lowering cholesterol, and seems specific to this class of drugs," Ms. Poynter said.

After adjusting for potential confounders including use of aspirin or NSAIDs for at least 5 years, Ashkenazi ethnicity, family history of colorectal cancer in a first-degree relative, participation in a sports activity, vegetable consumption, and hypercholesterolemia, in a matched analysis using conditional logistic regression, the odds ratio was 0.54 for statin use, for a 46% reduction in the risk of colorectal cancer (95% CI 0.39 to 0.75). "This suggests that the protective association that we observed between statins and colorectal cancer is not explained by other measured factors," she said.

When the data were analyzed separately for colon and rectal cancer, significant protective effects were seen for both cancers (OR = 0.53 for colon cancer and OR = 0.38 for rectal cancer).

At an ASCO press conference, Dr. Gruber said, "The data are very exciting for future clinical trials. There is a compelling rationale to investigate statins in other cancers besides colon cancer."