F. Turati, C. Galeone, M. Rota, C. Pelucchi, E. Negri, V. Bagnardi, G. Corrao, P. Boffetta, C. La Vecchia. Alcohol and liver cancer: a systematic review and meta-analysis of prospective studies. Annals of Oncology 2014; advance publication; doi:10.1093/annonc/mdu020
Despite several studies support a positive association between heavy alcohol consumption and liver cancer risk, a consistent dose–risk relationship has not yet been established. We carried out a systematic review and a meta-analysis of the association between alcohol intake and liver cancer occurrence, following the Meta-analysis Of Observational Studies in Epidemiology guidelines. We searched for cohort and nested case–control studies on the general population published before April 2013, using PubMed and EMBASE. Summary meta-analytic relative risks (RRs) were estimated using random-effect models. We included 16 articles (19 cohorts) for a total of 4445 incident cases and 5550 deaths from liver cancer.
Compared with non-drinking, the pooled RRs were 0.91 (95% confidence interval, CI, 0.81–1.02) for moderate drinking (<3 drinks per day) and 1.16 (95% CI, 1.01–1.34) for heavy drinking (≥3 drinks per day), with significant heterogeneity among studies. The dose–risk curve suggested a linear relationship with increasing alcohol intake in drinkers, with estimated excess risk of 46% for 50 g of ethanol per day and 66% for 100 g per day.
This systematic review suggests a moderate detrimental role of consumption of 3 or more alcoholic drinks per day on liver cancer, and a lack of association with moderate drinking. Our results have to be taken with due caution on account of the possible limitations of the original studies included in the meta-analysis.
Hepatic cirrhosis frequently precedes the development of liver cancer, and excessive alcohol consumption is known to be a cause of cirrhosis. A meta-analysis by Bagnardi et al (2001), based primarily on case-control studies, revealed an increase in liver cancer risk even among subjects reporting 25 g/day (about 2 typical drinks) of alcohol. A more recent study by Bagnardi et al (2013) of 7 cohort and 13 case-control studies revealed no association between light drinking (up to about 1 typical drink/day) and cancer of the liver (RR = 1.03, 95% CI 0.90–1.17).
The investigators in the present study carried out a meta-analysis to evaluate the association of alcohol consumption to liver cancer, using data from 19 cohorts with a large number of cases: 4,445 incident cases and 5,550 deaths from liver cancer. Their analysis indicates that liver cancer is related to heavy alcohol intake, but not to light-to-moderate drinking. The authors conclude: “This systematic review suggests a moderate detrimental role of consumption of 3 or more alcoholic drinks per day on liver cancer, and a lack of association with moderate drinking.”
Specific comments on the study: Forum reviewers considered this to be a well-done study, using appropriate statistical methodology. Only prospective studies were included in the analysis, with alcohol intake data collected at baseline from presumably healthy subjects. Most cohorts were from Asia (including one from Hawaii based on subjects of Japanese ancestry), with only 3 based on other Americans or Europeans.
It was noted that for moderate drinking (defined in the study as < 3 drinks/day), none of the cohort or nested case-control studies showed a statistically significant increase in risk of liver cancer, and the overall estimated RR was 0.91 (CI 0.81, 1.02). This is strong evidence that moderate drinking does not increase the risk of liver cancer. For subjects reporting 3 or more drinks/day, there was an overall estimated risk of 1.16 (CI 1.01, 1.34). A clearer increase in the relative risk of cancer was seen if higher cut points were used to define heavy drinking.
There were very similar effects for men and women, but there were differences noted according to geography: generally, there was little effect of alcohol among studies from Asia but a reduction in risk among subjects in European and US studies. When adjustments were made for previous hepatitis, there were no effects seen for moderate drinking, both for those with and those without such a history.
Reviewer Ellison commented that only 8 of the studies adjusted for body mass index, and the authors did not specifically comment on potential cirrhosis/cancer associated with non-alcoholic fatty liver disease or NAFLD. NALFD is now the most common type of liver disease in the US, and some studies show that moderate drinking decreases the risk of this disease.
Reviewer Finkel pointed out how complex the etiology of liver cancer is. “Liver cancer is a complicated subject: alcohol, hepatitis viruses, iron overload, other toxins, parasites, cirrhosis, and fatty liver, at least, enter the picture. So does the notable finding of several published studies that demonstrate moderate alcohol consumption is associated with beneficial effects among people with non-alcoholic fatty liver disease. I find this paper acceptable as far as it can go, but meta-analyses of such complexities have considerable limitations.”
Reviewer Lanzmann-Petithory commented that this study does not assess risk according to type of alcoholic beverage. “There are quite a lot of papers about protective effects of resveratrol (and other polyphenols) against hepatoma in cell or animal models. It could be that wine has not the same relation with liver cancer as other alcoholic beverages. For example, in our Nancy COLOR study, we observed a strong dose-dependent relation with other types of alcohol and a non-significant trend of protection with wine preference. It is to note that fruits are considered as protective against liver cancer so it would be not surprising if wine, this fermented fruit juice, has not the same relation with liver cancer risk. Differences by type of beverage, not evaluated in the present study, could explain partly the low correlation and even protective effect at moderate intake among certain [geographically defined] cohorts (given that wine is the most common alcoholic beverage in Europe but not in Asia).”
Forum member Skovenborg had several remarks about this study. “This seems to be a well done review with some caveats:
(1) Significant heterogeneity among studies;
(2) No stratification for BMI or smoking;
(3) No information about drinking pattern.”
In terms of only having baseline assessments of alcohol intake in this analysis, Skovenborg continues: “It is interesting that the importance of whether some participants stop drinking during the observation period is of less significance concerning cancer because clear evidence of decreases in the risk of cancer only appear after 20 years of abstinence.”
Reviewer Goldfinger stated: “I agree with others that this is a well-done study. While it would have been preferable to have had more than two levels of consumption, the favorable finding associated with moderate intake is notable. Further, I believe that reporting BMI among subjects would be interesting, as obesity predisposes to NAFLD and, among such subjects, the risk of hepatic fatty infiltration (arguably a precursor to primary liver malignancy) is reduced with moderate drinking.”
Forum member Teissedre had questions about the overall relation between alcohol consumption and cirrhosis. Ellison considered that there is no question that high levels of alcohol intake over time, especially among alcoholics, may lead to liver cirrhosis, which increases the risk of liver cancer. The key question relates to a potential increase in risk from light-to-moderate drinking. The present study suggests no increase in the risk of liver cancer from such levels of alcohol consumption. The recent meta-analysis by Bagnardi et al (2013) similarly showed that while they found an increased risk of liver cancer for heavy drinkers, they found no increase from light drinking, which they defined as up to about one drink/day. Bagnardi et al concluded: “Given the association between heavy alcoholic beverage consumption and liver cancer, our results suggest the existence of a threshold dose below which the effect of alcohol is negligible.”
References from Forum review
Bagnardi V, Blangiardo M, La Vecchia C, Corrao G. A meta-analysis of alcohol drinking and cancer risk. Br J Cancer 2001;85:1700-1705.
Bagnardi V, Rota M, Botteri E, et al. Light alcohol drinking and cancer: a meta-analysis. Ann Oncol 201324:301-308. doi: 10.1093/annonc/mds337.
Hepatic cirrhosis frequently precedes the development of liver cancer, and excessive alcohol consumption is known to be one cause of cirrhosis. The investigators in the present study carried out a meta-analysis to evaluate the association of alcohol consumption with liver cancer. They used data from 19 prospectively studied cohorts with a large total number of cases: 4,445 incident cases and 5,550 deaths from liver cancer.
The analysis from this large study indicates that liver cancer is related to heavier alcohol intake, but not to light-to-moderate drinking (with the latter defined in this study as < 3 typical drinks per day). The authors conclude: “This systematic review suggests a moderate detrimental role of consumption of 3 or more alcoholic drinks per day on liver cancer, and a lack of association with moderate drinking.”
Forum reviewers considered this to be a well-done analysis with appropriate statistical methodology. While the authors were unable to test the potential effects of pattern of drinking (regular moderate versus binge drinking), the type of beverage (wine or other beverages), or potential effects of obesity (with the latter being a key factor associated with the most common type of liver disease in the US, non-alcoholic fatty liver disease), the results of this study support most other research indicating that more than moderate drinking increases the risk of cirrhosis and liver cancer. However, the findings of no effect from what the authors considered “moderate” drinking (< 3 drinks/day, which exceeds the guidelines for most countries), also fits with another recent meta-analysis showing no increase in risk of liver cancer to be associated with light drinking.
There are many adverse health effects of heavy alcohol consumption. Liver cirrhosis, which frequently precedes the development of liver cancer, is one such potential outcome. On the other hand, this study suggests that moderate alcohol intake does not increase the risk of liver cancer.
* * *
Comments on this critique by the International Scientific Forum on Alcohol Research were provided by the following members:
Erik Skovenborg, MD, Scandinavian Medical Alcohol Board, Practitioner, Aarhus, Denmark
Harvey Finkel, MD, Hematology/Oncology, Boston University Medical Center, Boston, MA, USA
Creina Stockley, PhD, MBA, Clinical Pharmacology, Health and Regulatory Information Manager, Australian Wine Research Institute, Glen Osmond, South Australia, Australia
Tedd Goldfinger, DO, FACC, Desert Cardiology of Tucson Heart Center, Dept. of Cardiology, University of Arizona School of Medicine, Tucson, Arizona, USA
Dominique Lanzmann-Petithory,MD, PhD, Nutrition/Cardiology, Praticien Hospitalier Hôpital Emile Roux, Paris, France
Pierre-Louis Teissedre, PhD, Faculty of Oenology – ISVV, University Victor Segalen Bordeaux 2, Bordeaux, France
R. Curtis Ellison, MD, Section of Preventive Medicine & Epidemiology, Boston University School of Medicine, Boston, MA, USA