Is heavy alcohol use associated with young-onset pancreatic cancer?

Summary:

In the article that accompanies this editorial the authors demonstrate that compared to non-drinkers, alcohol consumption >2 drinks per day for men and >1 drink per day for women is prospectively associated with a weak but statistically significant elevated risk for total pancreatic cancer among individuals diagnosed with pancreatic cancer at ages < 50 years old. This research adds to the evidence that alcohol is associated with pancreatic cancer risk and supports implementation of the current public health recommendations that limit alcohol consumption for cancer prevention.

Total pancreatic cancer incidence has been increasing in the United States (US) and globally during the past 20 years.^1,2 In the US, its incidence from 2010 through 2019 increased across all age groups including those < 50 years old.³ This increase in a relatively short time-frame suggests that exogenous factors may be contributing. More than 85% of pancreatic cancer is pancreatic ductal adenocarcinoma.⁴ History of diabetes and chronic pancreatitis, tobacco use, and being overweight or obese are established, potentially modifiable risk factors for pancreatic ductal adenocarcinoma.^5–8 Pancreatic cancer is highly fatal, ranking third for cancer mortality the US⁹ and is projected to surpass colorectal cancer by 2040.¹⁰ There are no screening strategies for pancreatic cancer except for individuals with a genetic predisposition or family history at high risk;¹¹ therefore, it is typically diagnosed at advanced stages when treatment options are limited. Identifying modifiable risk factors offers an opportunity for prevention to reduce the burden of pancreatic cancer, particularly for young-onset disease.

Alcohol, particularly heavy alcohol use could plausibly promote pancreatic carcinogenesis through several hypothesized mechanisms.¹² For example, heavy alcohol use is the most common cause of both acute and chronic pancreatitis and could contribute to pancreatic cancer development via inflammation and pancreatitis, an established pancreatic cancer risk factor.¹² Alcohol metabolism generates reactive oxygen species, oxidative stress, and acetaldehyde production, a known carcinogen in humans¹³, which could potentially cause DNA damage and mutations in the pancreas.¹² Non-oxidative alcohol metabolism could also lead to pancreatic tissue injury in the presence of several enzymes from production of endogenous metabolites including fatty acid ethyl esters, phosphatidylethanol, ethyl glucuronide, and ethyl sulfate.¹²

Consensus reports suggest that the evidence for an association between alcohol use and pancreatic cancer risk is inconclusive.^13,14 A recent pooled study of 30 prospective cohorts of participants with a median baseline age of 57 years (10th, 90th percentiles 40, 69 years) and > 10,000 cases that was not part of recent reviews, showed significant weak to modest positive associations (HR, 1.12 to 1.36) between alcohol consumption and pancreatic ductal adenocarcinoma for greater than 1 drink per day (≥15 g/day) among women and 2 drinks per day (≥30 g/day) among men, compared to light drinkers (0.1–<5 g/day).¹⁵ Most previous research evaluating associations between alcohol use and pancreatic cancer were conducted in people older than 50 years.^15,16 To address this gap, in this issue of the Journal of Clinical Oncology, Park et. al.¹⁶ examined the association between modest to heavy alcohol intake and total pancreatic cancer within a large nationwide Korean cohort of 6.3 million men and women, aged 20 to 39 years who underwent a health screening between 2009 to 2012 that included questions about frequency of alcohol consumption. One standard drink in Korea has 8 g of ethanol, while in the US it contains 14 g of ethanol.¹⁶ For comparability to previous published studies, the authors define alcohol use into the following categories: nondrinkers, light-to-moderate drinkers (<30 g/day for men and <16 g/day for women), and heavy drinkers (≥30 g/day for men and ≥16 g/day for women). During a median follow-up up to 12 years, 1,515 confirmed incident pancreatic cancers were identified at ages <50 years. Compared to non-drinkers, the authors observed a marginally significant elevated pancreatic cancer risk (HR=1.19, 95%CI 1.004–1.42) with heavy alcohol use (≥16 grams per day among women and ≥ 30 grams per day among men) in multivariable adjusted sex combined models with the magnitude of the association being greater in men (HR, 1.29, 95% 1.05–1.58) than in women (HR, 0.88, 95% 0.61–1.27), although the interaction by sex was not significant (P-interaction=0.14). Furthermore, alcohol use frequency ≥3 times per week was associated with an increased risk (HR, 1.23; 95% CI, 1.01–1.51).

A limitation of the analysis is that there could be residual confounding, particularly for smoking from the variable used to adjust for smoking in the models and lack of information on secondhand smoke. Cigarette smoke is highly correlated with alcohol use and examining associations among never smokers helps to minimize residual confounding by smoking. Although smoking status did not significantly modify associations between heavy alcohol use and pancreatic cancer (P-interaction=0.22), significant risk associations were only observed among current smokers (HR, 1.36, 95% CI, 1.05, 1.76), not among non-smokers (HR, 1.13, 95% CI, 0.87–1.47) or never smokers (HR, 1.31, 95% CI, 0.96–1.78), however the magnitude of the association in never smokers was similar to that of current smokers. The prevalence of smoking in South Korea is 19% overall with the prevalence in men being 30% compared to 5% in women.¹⁷ It is possible that the lack of a positive alcohol use and pancreatic cancer association among women, could be related to very few women smoking and hence no residual confounding, although the prevalence of heavy drinking also was low among women. There may also be other unmeasured correlations with alcohol use related to lifestyle such as diet or other exposures that are not controlled in the analyses. The authors did not have genetic data on the aldehyde dehydrogenase 2 (ALDH2) gene polymorphisms or on alcohol flushing which has been estimated to be present in 40% of the Korean population.¹⁸ The potential effect of ALDH2 on cancer risk among those who carry the inactive genotype (ALDH2*2) is complex because it could lead to reduced alcohol use due to flushing or alternatively, among those who drink alcohol, accumulation of acetaldehyde, a known carcinogen,¹³ and increased cancer risk.¹⁹ The results from this study might not be generalizable to other populations, race-ethnic groups, or individuals not from East Asian ancestries. The authors did not evaluate cancer associations by different alcohol types (e.g. beer, liquor, wine). Finally, the outcome was total pancreatic cancer and did not distinguish pancreatic ductal adenocarcinoma from other pancreatic cancer tumor types or neuroendocrine tumors. Pancreatic neuroendocrine tumors occur more often among people aged <50 years,²⁰ while 90% of pancreatic ductal adenocarcinoma occurs among people age 55 years or older.⁹ It is uncertain what proportion of the young-onset pancreatic cancer in the analysis were neuroendocrine tumors or how inclusion of neuroendocrine tumors might have affected the alcohol risk estimates.

Despite these limitations, the authors demonstrate that heavy alcohol use is associated with young-onset pancreatic cancer in a large nationwide prospective cohort representative of the Korean population and with more than 1500 incident cases.¹⁶ The findings from Park et. al. adds to the evidence that alcohol is associated with pancreatic cancer risk and to the evidence supporting the current recommendations to either not drink or limit alcohol to 2 drinks or less per day for men and 1 drink or less per day for women, when alcohol is consumed.²¹ The findings from this study warrant further investigation and confirmation in other populations.