Drs. Mason and Levesque thoroughly review data from intervention trials and epidemiologic studies that suggest a role for folate in preventing cancer of the colorectum and, to a lesser degree, cancers of the uterine cervix, lung, esophagus, and stomach. The authors also provide a comprehensive discussion of the possible mechanisms by which folate may prevent cancer, in particular, the relationship between folate status and DNA methylation.
Drs. Mason and Levesque thoroughly review data from interventiontrials and epidemiologic studies that suggest a role for folatein preventing cancer of the colorectum and, to a lesser degree,cancers of the uterine cervix, lung, esophagus, and stomach. Theauthors also provide a comprehensive discussion of the possiblemechanisms by which folate may prevent cancer, in particular,the relationship between folate status and DNA methylation.
The article begins with a historical perspective that relatesthe development of megaloblastosis and possibly dysplasia resultingfrom folate deficiency. Perhaps the most interesting of the earlyaccounts linking folate deficiency to megaloblastic anemia isthe experiment by Herbert, in which he deprived himself of dietaryfolate. However, as Mason and Levesque indicate, the observationslinking megaloblastosis to dysplasia are a priori evidence. Althoughthere is a strong link between folate deficiency and hypomethylationand a link between hypomethylation of specific genes and cancer,the connection between folate-deficient diets and dysplasia iscircumstantial.
The notion that folate deficiency functions as a cocarcinogenis supported by the observation that an increase in cervical cancerwas evident only in folate-deficient patients who had a concurrenthuman papilloma virus-16 (HPV-16) infection. If this notion isaccurate and folate deficiency is manifested as dysplasia onlywhen specific organisms are present or in the context of a specificbiochemical environment, there is a very definite need to conductstudies to determine the exact nature of these corequirementsfor cancer.
How Much Folate? Is More Better?
The authors correctly state that results from epidemiologic studiesdo not establish a cause-and-effect relationship between folatestatus and cancer. To take this one step further, results fromstudies that depend on dietary recall or that do not standardizefood preparation do not take into account the fact that food preparationand cooking substantially influence the amount of folate actuallyingested.
The authors describe the results of two rat studies coauthoredby Dr. Mason (one of which is in press) that support the contentionthat increased dietary folate intake decreases the incidence ofcolorectal cancer. To our knowledge, these are the only studiesin which the relationship between colorectal cancer and folatestatus has been examined in animals. In rats, a folate-deficientdiet decreases the incidence of mammary cancer, but results froman older study suggests that folate causes regression of mammarytumors in mice.[3,4]
As alluded to by the authors, the cancer chemopreventive effectof folate may be specific to certain organs, but one cannot dismissthe possibility that increasing dietary consumption of folatemay have less beneficial or perhaps even harmful effects in sometissues or organs. It seems irrefutable that other dietary componentswill interact with folate to influence the development of dysplasiaand that this interaction may also be organ-specific.
A particularly interesting point made by the authors is that lowerred blood cell (RBC) folate levels that are within the conventionalnormal range are associated with a higher incidence of colorectaland uterine cancer than are higher folate levels. As discussedby McNulty, the currently recommended dietary allowances of folatein both the United Kingdom and the United States have been lowered,presumably in response to a lack of evidence that a higher intakeis necessary or beneficial. In view of the aforementioned results,further study is required to determine whether a reevaluationof the current dietary allowances may be needed.
With regard to gender, race, and socioeconomic status, it wouldbe informative to know whether folate or its metabolites havedifferent effects depending on the indicator. For example, Americanwhites consume significantly more folate than do African-Americans,but it is not known to what extent, if any, this accounts forthe observed racial differences in the incidence of various cancers.
Alcohol, Tobacco, and Folates
The interaction among alcohol, tobacco, and folate as it influencescolorectal cancer is an area in which there are more questionsthan answers. As the authors note, high alcohol consumption coupledwith a low folate diet translates into a fourfold higher incidenceof colorectal cancer. Collectively, the data supporting an associationbetween alcohol and colorectal cancer are unconvincing and aremost compelling with regard to beer drinking and rectal cancer.Interestingly, beer accounts for 10% of the total folate intakeby adults in the United Kingdom. Given that a small decrease inRBC folate levels appears to result in a higher incidence of colorectalcancer, encouraging temperance, at least with regard to beer,would lower folate levels, producing a net effect that would notnecessarily be beneficial!
As mentioned earlier, perhaps folate deficiency functions as acocarcinogen with alcohol in the colon. In one study, a methyl-deficient(low-methionine, low-folate) diet in conjunction with high alcoholintake has been associated with a higher risk of human colorectalcancer. In this study, smoking was not a confounding factor,and the most recent report focusing on tobacco and colorectalcancer does not support a correlation between the two. Althoughthe data are conflicting, the consensus appears to be that ifsmoking does increase the risk of colorectal cancer, it does soonly in those who smoke for more than 35 years, and the risk isrelatively greater for the rectum.
The authors state that chronic use of tobacco is associated withdecreased blood folate levels. Thus, owing to the associationbetween cancer incidence and folate concentration, it would standto reason that chronic smokers would have a higher incidence ofcolorectal cancer. This has not been observed, however. A hostof confounding issues enter into this equation, most of whichare just beginning to be approached experimentally or in clinicalstudies.
The authors detail various potential mechanisms by which folatemay prevent cancer. The role of folate in altering DNA methylationstatus is discussed most thoroughly. DNA methylation results fromtransfer of a methyl group from S-adenosylmethionine to deoxycytidinelocated primarily in CpG islands. The primary dietary sourcesof these methyl groups are folate, methionine, and choline.
In general, methylation is a mechanism for regulating gene activity.DNA hypomethylation is recognized as a very early event in adenomaformation, specifically in adenomas 0.5 cm.or less It is notcertain how methylation influences carcinogenesis, but it is thoughtthat a change in chromatin packing enhances the accessibilityof carcinogens to DNA. Alternatively, hypomethylation may leadto global genomic instability and, consequently, aberrant chromosomepairing and dysjunction during mitosis.
Interestingly, hypermethylation can also influence tumorigenesisby downregulating expression of important genes. Both hypomethylationand hypermethylation may be regional and may greatly influencethe activation state of important oncogenes or tumor-suppressorgenes. The methylation status of specific genes may therebydetermine the outcome of subsequent mutational events. It wouldappear that DNA methylation greatly influences tumor formation,and an important body of work indicates that diet greatly influencesmethylation status. This is a highly pertinent area of study thatwill continue to generate considerable interest.
There is a growing consensus that chemoprevention will be mostsuccessful through the use of multiple micronutrients. Unfortunately,as noted by Dr. Frank Meyskens at the most recent AACR meeting,we currently lack good intermediate biomarkers to assess the efficacyof dietary interventions in preventing cancer. DNA methylationstatus may be a good biomarker, but considerable work remainsto be done to establish its utility in chemoprevention studies.
1. Herbert V: Experimental nutritional folate deficiency in man.Trans Assoc Am Physicians 75:307-320, 1962.
2. McNulty H: Folate requirements for health in different populationgroups. Br J Biomed Sci 52:110-119, 1995.
3. Baggott JE, Vaughn WH, Juliana MM, et al: Effects of folatedeficiency and supplementation on methylnitrosurea-induced ratmammary tumors. J Natl Cancer Inst 84:1740-1744, 1992.
4. Lewisohn R, Leuchtenberger C, Leuchtenberger R, et al: Theinfluence of liver L. casei factor on spontaneous breast cancerin mice. Science 104:436-437, 1946.
5. Giovannucci E, Rimm EB, Ascherio A, et al: Alcohol, low-methionine-low-folatediets, and risk or colon cancer in men. J Natl Cancer Inst 87:265-273,1995.
6. Nyren O, Bergstrom R, Nystrom L, et al: Smoking and colorectalcancer: A 20-year follow-up study of Swedish construction workers.J Natl Cancer Inst 88:1302-1307, 1996.
7. Jennings E: Folic acid as a cancer-preventing agent. Med Hypotheses45:297-303, 1995.
8. Nivatvongs S, Dorudi S: Colorectal polyps and their management,in NS Williams (ed): Colorectal Cancer, pp 39-54. New York, ChurchillLivingstone, 1996.
9. Baylin SB, Makos M, Wu J, et al: Abnormal patterns of DNA methylationin human neoplasia: Potential consequences for tumor progression.Cancer Cells 3:383-390, 1991.
10. Is chemoprevention overrated or underfunded? (commentary).J Natl Cancer Inst 88:947-949, 1996.