Statins Cut Colon Ca Risk, Retrospective Study Shows; Further Investigation Warranted

NEW ORLEANS-In a retrospective,observational study ofnearly 4,000 patients in northern Israel,statin use for at least 5 years reducedthe risk of colorectal cancer by46%, after adjustment for known protectivefactors, including use of aspirinand nonsteroidal anti-inflammatorydrugs (NSAIDs). The effects seendo not appear to be the result of areduction in cholesterol levels. Whilethe investigators, from the MolecularEpidemiology of Colorectal Cancer(MECC ) Study Group, cautioned thatit is premature to change either thestandard of care for colorectal canceror the indications for statins, they saidstatins merit further investigation inthis area.MECC StudyJenny N. Poynter, MPH, of the Universityof Michigan, presented thestudy results at the 40th Annual Meetingof the Society of Clinical Oncology(abstract 1), on behalf of the MECCStudy Group. MECC is led by StephenB. Gruber, MD, PhD, MPH, directorof clinical cancer genetics, Universityof Michigan Comprehensive CancerCenter, and Gadi Rennert, MD, directorof the National Cancer ControlCenter for Clalit Health Services, Israel'slargest HMO.The MECC study, initiated in 1999with a $4.8 million grant from theNational Cancer Institute and additionalfunding from the Irving Weinsteinand Ravitz Foundations, exploreshow genes, diet, exercise, and otherfactors interact to produce colon cancerin individuals.Given that the descendants of AshkenaziJews have unusually high ratesof colon cancer (with a novel cancersusceptibility allele, APC I1307K, identifiedin 6% and appearing to double colorectal cancer risk), the researchersbelieved an Israeli-based populationstudy including this ethnic group, aswell as Arab patients and Jews of non-Ashkenazi ethnicity, might provideinteresting information about possiblenongenetic factors influencing developmentof this disease.Statins were assessed in this population,Ms. Poynter said, "because theyinhibit HMG CoA reductase, which isoverexpressed in colorectal cancer celllines, and induce apoptosis in colorectalcancer cell lines at physiologicallyrelevant concentrations. They havealso been shown to reduce tumor formationin animal models of colorectalcancer."Eligible patients were those individualsdiagnosed with colon or rectalcancer between 1998 and 2004 (n =1,814 cases). Controls (n = 1,959) werematched to cases based on age, sex,and Jewish vs non-Jewish ethnicity.The researchers also performedmatched analyses on 1,570 perfectlymatched case-control pairs.Reduced Cancer Prevalancein 'Statin Users'Medication use, including statins,was assessed in a structured in-personinterview. Classified as "statin users"were those reporting use of any statinfor at least 5 years; nonusers had noreported statin use. Regular statin use(prescriptions filled more than threetimes per year) was validated fromprescription records from the ClalitHealth System (Israel)database for themajority of patients who reported usingthese drugs. All pathology was independentlyconfirmed by a single pathologistat the University of Michigan.Use of statins for at least 5 years wasreported by 328 participants: 106 cases(5.8%) and 222 controls (11.3%)."In an unadjusted, unmatched analysis,we saw a strong protective effect ofstatins, with an odds ratio (OR) of0.49, and a 95% confidence intervalfrom 0.38 to 0.62, and a P value thatwas highly significant [P < .0001],"Ms. Poynter said. "Similar results wereseen in a matched analysis of the 1,570pairs currently available."Pravastatin (Pravachol) and simvastatin(Zocor) were the two mostcommon statins used, accounting for44% and 52% of use, respectively. Bothdrugs were associated with a reducedrisk of colorectal cancer (OR = 0.45for pravastatin, 0.47 for simvastatin).The only other cholesterol-loweringdrug commonly used in this studywas the fibric acid derivative bezafibrate(Bezalip), which has pharmacologicactivity distinct from the statinsand lowers cholesterol by increasingactivity of lipoprotein lipase. In thestudy, 39 patients reported using bezafibratefor at least 5 years. No significantassociation was seen betweenbezafibrate use and incidence of colorectalcancer (OR = 1). "Thus, theprotective effect of statins is not likelyto be due to a general effect of loweringcholesterol, and seems specific tothis class of drugs," Ms. Poynter said.After adjusting for potential confoundersincluding use of aspirin orNSAIDs for at least 5 years, Ashkenaziethnicity, family history of colorectalcancer in a first-degree relative, participationin a sports activity, vegetableconsumption, and hypercholes-terolemia, in a matched analysis usingconditional logistic regression, theodds ratio was 0.54 for statin use, fora 46% reduction in the risk of colorectalcancer (95% CI 0.39 to 0.75). "Thissuggests that the protective associationthat we observed between statinsand colorectal cancer is not explainedby other measured factors," she said.When the data were analyzed separatelyfor colon and rectal cancers,significant protective effects were seenfor both of these cancers (OR = 0.53for colon cancer and OR = 0.38 forrectal cancer).At an ASCO press conference, todiscuss the findings, Dr. Gruber said,"The data are very exciting for futureclinical trials. There is a compellingrationale to investigate statins in othercancers besides colon cancer."