Breast Cancer

Key Question 1. What are the likely causal mechanisms by which alcohol contributes to the development of breast cancer? Which of the possible mechanisms (e.g., induction of P450 cytochromes and carcinogen metabolism, effects on blood hormone concentrations, effect of acetaldehyde or other alcohol metabolite on apoptosis and DNA repair, interactive effects on other nutritional factors, or others) are likely to be most important in breast cancer development?

Other Potential Mechanisms

A single human study by Hartman et al.⁷⁶ reported on increased level of biomarkers of oxidative stress such as α-tocopherol and isoprostane after alcohol consumption (see Table C-1 in Appendix C). Our searches identified five epidemiology studies^77–81 that also postulated a connection between biomarkers of inflammation, alcohol intake, and risk of breast cancer. Increased levels of biomarkers such as malondialdehyde,^77,79 isoprostanes,⁸¹ and catalase activity^78,80 were reported. We did not identify any experimental studies using animal or cell line models that examined other potential mechanisms.

Key Question 2. For the most likely mechanisms of action involving alcohol and the development of breast cancer, how might other factors modify the effect of alcohol on breast cancer (for example, age, latency of effect, intensity, duration, and recency of exposure, presence of co-carcinogens, presence of threshold effect)? Do the causal mechanisms vary by cell type or other tumor characteristics?

For this Key Question, we looked for studies that evaluated factors that modify the association of alcohol with biomarkers of risk of breast cancer. The human studies of alcohol consumption and hormone changes were performed in pre- and postmenopausal women but an actual age effect was not examined in these studies. The duration of consumption was relatively short; long term effects could not be calculated in these studies. However, we did identify one human study that examined biomarkers of oxidative stress and risk of carcinogenesis. Hartman et al. reported that in postmenopausal women who consumed 30 g alcohol per day, α-tocopherol decreased by 4.6% and isoprostane levels increased by 4.9%.⁷⁶ This study provides a possible link between oxidative stress and risk of breast cancer formation.

Table 7 and Table 8 contain an overview of the breast cancer studies included in this report in terms of study design and reporting issues that determined whether the study provides evidence of a direct or an indirect association between alcohol consumption and breast cancer. Route of administration, rate of absorption and metabolism, formulation and quantity of ethanol, and timing of the intervention, however, may reduce the generalizability of animal studies to a clinical setting. Although we evaluated cell line studies as part of our overall evidence evaluation, we did not include them in this table given that events such as confounding exposure, control for other risk factors, and cancer formation are not applicable to this model.

Table 7

Overall results from human breast cancer studies.

Table 8

Overall results from animal breast cancer studies.

Colorectal Cancer

Key Question 3. What are the likely causal mechanisms by which alcohol contributes to the development of colorectal cancer? Which of the possible mechanisms (e.g., induction of P450 cytochromes and carcinogen metabolism, effects on blood hormone concentrations, effect of acetaldehyde or other alcohol metabolite on apoptosis and DNA repair, interactive effects on other nutritional factors, or others) are likely to be most important in colorectal cancer development?

Acetaldehyde production in the colon. Exposure of colon mucosa to acetaldehyde from microbial metabolism of ethanol has been postulated as a mechanism for increasing the risk of developing colorectal cancer in two epidemiology studies (see Table 4)^55,56 and three experimental studies (see Table 5).^57,307,309 According to study authors, individual variations in human colonic flora may contribute to the risk of alcohol-related colorectal cancer,⁵⁵ and increased activity of intracolonic bacterial alcohol dehydrogenase may also play a role in increasing cancer risk.^{56,57,307,309}

Experimental human studies examining this subject are few (see Table C-4 in Appendix C). A study by Basuroy et al.⁸³ suggests that acetaldehyde disrupts epithelial tight junction and cell adhesion and through this mechanism increases the risk of colon cancer. Several animal studies also looked at the effects of acetaldehyde in the colon (see Table C-5 in Appendix C). These studies showed mucosal damage after ethanol consumption,¹²⁰ increased degradation of folate,⁷⁰ stimulation of rectal carcinogenesis,¹²² and an increased effect of carcinogens in the presence of acetaldehyde.¹²³ In cell line studies acetaldehyde exposure was reported to influence the initial steps of colonic carcinogenesis and later tumor development¹³⁶ and decrease the activity of some brush border enzymes.¹³⁷ Finally, a study using animal and cell line tissue found evidence that acetaldehyde stimulates cell proliferation in animal intestinal crypt cells and therefore acetaldehyde may act as a cocarcinogen in the colon.¹⁴¹ These studies suggest that acetaldehyde production in the colon may provide a potential causal mechanism by which alcohol contributes to the development of colon cancer.

Cell proliferation. Hyperproliferation of rectal mucosa after exposure to alcohol was postulated as a mechanism for increasing the risk of developing colorectal cancer in an epidemiology study by Simanowski et al.²³⁵ The authors examined rectal biopsies for proliferation markers such as histone H3 and Ki67 in 44 heavy drinkers and 26 controls. Heavy drinkers showed an increase in cell proliferation markers in the rectal mucosa compared to controls.²³⁵

An effect of ethanol consumption on cell proliferation in the colon was investigated in both animal and cell line studies in our primary evidence base. Several animal studies reported enhanced growth of mucosal tissue after chronic ethanol consumption.^125–127 Cell studies indicated that exposure to ethanol and acetaldehyde increases cell proliferation^74,136 and damages DNA which may contribute to cancer development.¹³⁵ Together these studies suggest that ethanol and acetaldehyde exposure in the colorectal mucosa may increase cell proliferation and be a potential mechanism connecting alcohol consumption to colorectal cancer risk.

DNA repair polymorphism. We identified 52 epidemiology studies that assessed DNA repair polymorphism and alcohol consumption. The majority of these studies suggested that DNA repair polymorphism may influence the risk of colorectal cancer.

Enzyme polymorphism. We identified 19 studies that assessed enzyme polymorphism in epidemiology studies: 13 examined alcohol and acetaldehyde dehydrogenase polymorphism;^19,241–252 five examined cytochrome P450 polymorphism;^236–240 and six examined methylenetetrahydrofolate reductase polymorphism.^{189,259–263} The majority of these studies reported enzyme polymorphism as a risk marker for colorectal cancer following moderate alcohol consumption.

Other potential mechanisms. Ethanol may also influence carcinogenesis in the colon and rectum through an interaction with carcinogens. Animal studies suggest that ethanol exposure in the colon increases the chances of tumor development,¹³² but other studies found no association between ethanol ingestion and colorectal carcinogenesis or instead reported inhibition of tumorigenesis.^73,130,131 Other possible mechanisms reported in animal studies include alcohol’s inhibition of folate metabolism⁷⁰ and DNA hypomethylation.¹²⁴

Key Question 4. For the most likely mechanisms of action involving alcohol and the development of colorectal cancer, how might other factors modify the effect of alcohol on colorectal cancer (for example, age, latency of effect, intensity, duration, and recency of exposure, presence of co-carcinogens, presence of threshold effect)? Do the causal mechanisms vary by cell type or other tumor characteristics?

For this Key Question, we looked for studies that evaluated factors that modify the association of alcohol with biomarkers of colorectal cancer risk. Few studies are available that examined factors that modify the effects of ethanol consumption on the risk of developing colorectal cancer. The study in human subjects in which biopsy samples were examined for damage after exposure to acetaldehyde did not report the influence of personal factors on the degree of damage generated.⁸³

Table 9 and Table 10 contain an overview of the colorectal cancer studies included in this report in terms of study design and reporting issues that determined whether the study provides evidence of a direct or an indirect association between alcohol consumption and colorectal cancer. Route of administration, rate of absorption and metabolism, formulation and quantity of ethanol, and timing of the intervention however may reduce the generalizability of animal studies to a clinical setting. Although we evaluated cell line studies as part of our overall evidence evaluation, we did not include them in this table given that events such as confounding exposure, control for other risk factors, and cancer formation are not applicable to this model.

Table 9

Overall results from human colorectal cancer study.

Table 10

Overall results from animal colorectal cancer studies.

Excluded Studies

Because this is a systematic review using specific inclusion and exclusion criteria with the creation of specific Key Questions, the report is directed at evidence that addresses each Key Question. None of the excluded studies (see Table D-1 in Appendix D) were left out for quality, design, conduct, integrity, or inaccuracy but rather because they did not address these Key Questions.

Future Research Goals

Our examination of the epidemiology literature correlating alcohol consumption with cancer risk has suggested many areas in which experimental research may provide insight into the actual mechanisms connecting cancer risk and alcohol consumption. For breast cancer these potential mechanisms are changes in circulating hormone levels and changes in hormone receptors, DNA-adduct formation, and various enzyme polymorphisms related to alcohol metabolism. For colorectal cancer these areas are DNA repair polymorphisms, mucosal cell proliferation, and various enzyme polymorphisms related to alcohol metabolism. Experimental studies in humans, animals, or cell lines have provided basic information on some but not all of these potential mechanisms.

The connection between alcohol intake and changes in estrogen levels and breast cancer risk has been studied in human, animal, and cell line studies. Future research in this area would seem to be warranted to determine the exact level of risk imposed by this pathway. A connection between cell proliferation and tumor progression in breast cancer has been suggested by animal studies but not in human studies and human-based studies in this area would seem to be warranted. Enzyme polymorphism in ethanol metabolism as well as in other metabolic pathways that may be influenced by alcohol may require more human-based studies as opposed to animal studies where polymorphism is not a factor. DNA adduct formation has not been well studied in human or animal studies and research in this area should be expanded. Oxidative stress and inflammation associated with alcohol consumption have been postulated as risk factors in epidemiology studies but not studied to any extent in hypothesis-generating studies. Oxidative stress and inflammation should be examined with better experimentally controlled studies.

Experimental human studies examining the connection between alcohol intake and colorectal cancer are few. Many potential mechanisms related to acetaldehyde production in the colon, cell proliferation due to ethanol or acetaldehyde exposure, alterations in DNA repair mechanisms, and the influence of carcinogens and alcohol in the colon need to be examined in human-based studies. Animal studies are also needed to examine the influence of bacterial flora, the effects of ethanol and acetaldehyde on the colon, especially changes in cell proliferation and DNA, and the interaction between carcinogens and ethanol and acetaldehyde.

Conclusions

Based on our systematic review of the literature, many potential mechanisms by which alcohol may influence the development of breast or colorectal cancers have been explored but the exact connection or connections remain unclear. The evidence points in several directions but the importance of any one mechanism is not apparent at this time.

Table 11 through Table 13 summarizes the mechanisms on alcohol consumption and the risk of breast cancer as presented in studies identified in this report. Six human, five animal and five cell line studies reported on changes in hormonal levels as the potential causal mechanism by which alcohol consumption may contribute to the development of breast cancer. Our findings are comparable to the most commonly reported mechanisms in most of the breast cancer epidemiology studies summarized in Table 3.

Table 11

Reported mechanisms in human breast cancer studies.

Table 12

Reported mechanisms in animal breast cancer studies.

Table 13

Reported mechanisms in cell line breast cancer studies.

Table 14 through Table 17 summarizes the mechanisms of alcohol consumption and the risk of colorectal cancer as presented in studies identified in this report. One human study exposed colonic mucosa biopsies to vapor-phase acetaldehyde and reported an effect of acetaldehyde on cell adhesion as the most likely causal mechanisms by which alcohol consumption may contribute to the development of colorectal cancer. In contrast, nine animal studies reported a local toxic effect of acetaldehyde resulting in mucosal damage as the most likely causal mechanism by which alcohol consumption may contribute to development of colorectal cancer. Other mechanisms identified in this report include:

Table 14

Reported mechanisms in human colorectal cancer study.

Table 15

Reported mechanisms in animal colorectal cancer studies.

Table 16

Reported mechanisms in cell line colorectal cancer studies.

Table 17

Reported mechanisms in combination (animal, cell lines) colorectal cancer study.

• increase in cytochrome P4502E1 expression (two animal studies)
• effect on DNA synthesis and methylation (two animal studies, two cell line studies)
• effect on cell proliferation (two cell line studies)
• apoptotic cell death (three cell line studies)
• effect on various stages of carcinogenesis (two animal studies)
• changes in polyamine content (one animal study)
• effect of acetaldehyde on brush border enzymes (one cell line study)
• modulation of gene expression (one cell line study).

Our findings are comparable to some of the most common mechanisms (e.g., colonic microbial production of acetaldehyde, effect on DNA methylation and synthesis) reported by the colorectal cancer epidemiology studies summarized in Table 4.

Limitations

The evidence base for the report included 66 studies:

• six human studies (five breast cancer, one colorectal cancer)
• 34 animal studies (15 breast cancer, 19 colorectal cancer)
• 25 cell line studies (15 breast cancer, 10 colorectal cancer)
• one combination study (animal, cell line) on colorectal cancer.

Therefore the evidence in support of any potential mechanism connecting alcohol intake to cancer development is based largely on animal models. Animal models are important tools for understanding disease mechanisms but they have limitations when predicting the actual course of events in humans.⁸² Reviews of animal studies have shown that there is a tendency to publish studies with positive results and not to publish studies that suggest no difference in measured outcomes (i.e., publication bias). Therefore studies that could possibly rule out mechanisms connecting alcohol and cancer may not be published. Positive results in animal studies may not translate to a clinical setting because carcinogens were administered in a controlled setting that is not characteristic of human conditions. Most experimental animals are young and rarely have comorbidities, a situation that may also limit generalizability of animal studies to clinical studies.³¹²

Few human studies met the inclusion criteria for this report and this limited the comparisons that could be made between the findings of animal studies and those in human studies. Exact alcohol exposure can be controlled in animal studies but few human studies have done the same. While the four breast cancer human studies actually administered and quantified the amount of ethanol, the only colorectal cancer study administered acetaldehyde to biopsied colonic mucosa. Because of the limited number of human studies in our evidence base, we did look at potential mechanisms suggested in epidemiology studies and compared them to mechanisms examined in animal and cell line studies.