Biliary Tract Cancers Pooling Project

Subproject

Liver Cancer Pooling Project (LCPP)

Peter Campbell-peter.campbell@einsteinmed.org

NIH Intramural funding

We propose to expand the Cohort Consortium Liver Cancer Pooling Project (LCPP) to include biliary tract cancers (i.e., cancers of the gallbladder, extrahepatic bile duct, ampulla of Vater, and overlapping lesions of the biliary tract). As the Cohorts involved in the LCPP have already contributed most variables relevant to biliary tract cancer, we anticipate that this expansion would require minimal effort. For example, the LCPP will examine the associations of smoking and diabetes with liver cancer. Very similar analyses of biliary tract cancer could also be conducted. At the present time, one other effort in a biliary tract cancer is being undertaken in the Cohort Consortium: the Obesity and Rare Cancers Pooling Project is planning to examine the relationship between BMI and gallbladder cancer. That analysis is being led by Peter Campbell, who will be a co-leader of this new, more comprehensive, Biliary Tract Cancer effort.

Gallbladder cancer is the most common type of biliary tract cancer, accounting for about 50% of biliary tract cancers. Cancers of the extrahepatic bile duct and ampulla of Vater display distinct demographic patterns and molecular profiles, indicating that they are separate disease entities (1). For example, gallbladder cancer occurs about two-to-five fold more often in women than in men (2, 3), while other biliary tract cancers tend to be slightly more common in men (3). Given the rarity and exceptionally poor prognosis of these cancers [e.g., fewer than 10000 total incident cases are expected in the US in 2012 (4)], little is known about their etiology beyond a positive association with history of cholesterol gallstones, which increases risk of gallbladder cancer by about four-to-24-fold (5, 6) and also increases the risk of other biliary tract cancers (1). The Cohort Consortium is unique in its ability to prospectively evaluate risk factors for individual types of biliary tract cancers by combining across studies. A key advantage of this resource is that most of the necessary data have been already collected and harmonized, meaning that only a few additional variables would be required from each participating cohort in order to conduct this project. Based on variables that are currently available from the LCPP, as well as variables that are likely to be available in many cohorts, we propose the following analyses:

BMI study expanded to all biliary tract cancers
The current analysis of BMI and gallbladder cancer could be greatly enhanced by adding the other biliary tract cancers (i.e., extrahepatic bile duct, ampulla of Vater, and overlapping lesions of the biliary tract).

Smoking
Smoking has been associated with an increased risk of gallbladder and other biliary tract cancers in some studies, but the data are inconsistent (7). A pooled effort of prospective studies, as proposed here, offers a timely opportunity to evaluate the association between smoking and biliary tract cancers.

Diabetes
A recent meta-analysis reported similar summary risk estimates for the associations of diabetes with cancers of the gallbladder (summary RR: 1.52, 95% CI: 1.26–1.84), extrahepatic bile duct (summary RR: 1.59, 95% CI: 1.26–1.99), and ampulla of Vater (summary RR: 1.60, 95% CI: 0.60–4.22). (8) However, only three cohort studies have reported on diabetes and cancer of the extrahepatic bile duct, and only one cohort has published on diabetes and ampulla of Vater cancer, limiting the stability of prospective estimates. The pooled analysis proposed here would also allow for a more detailed evaluation of the association between diabetes and biliary tract cancer over time and it could evaluate the potential modifying or confounding impact of gallstones on the association between diabetes and biliary tract cancers.

Anti-inflammatory drugs
It has been suggested that inflammation, particularly inflammatory conditions related to gallstones, plays an important role in biliary tract carcinogenesis. Inflammation in the biliary tract may lead to dysplastic and neoplastic changes. Examination of cholecystectomy specimens removed during treatment for gallstones or cholecystitis suggests a sequence of precursor epithelial changes evolving through hyperplasia, atypical hyperplasia, metaplasia, dysplasia, and carcinoma in situ. Regardless of the presence or absence of gallstones, dysplastic and neoplastic changes in the gallbladder and bile duct have been associated with a variety of inflammatory conditions, including cholecystitis, primary sclerosing cholangitis, fibrotic reactions, and strictures (9). The precise mechanisms linking inflammation and carcinogenesis are not entirely clear but may involve nitric oxide production in inflamed tissue leading to oxidative stress and DNA damage (10). Case-control data suggests that nonsteroidal anti-inflammatory drugs (NSAIDs) are associated with reduced risk of gallbladder cancer (11). Prospective evaluation in the Cohort Consortium is warranted.

Family History of Cancer
In a recent case control study, family history of liver cancer was associated with an increased risk of extrahepatic bile duct cancer but not gallbladder or ampulla of Vater cancer (12). Family history of any cancer was less common among gallbladder cancer patients compared to population-based controls (13). Given that gallbladder cancer has been hypothesized to be hormonally-related, family history of breast cancer and family history of prostate cancer might also be associated with higher risk. By pooling biliary tract cancer cases in the Cohort Consortium, it would be possible to evaluate these associations.

The purpose of this proposal is to prospectively evaluate risk factors for biliary tract cancer. Given the rarity of cancers of the gallbladder, extrahepatic bile ducts, and ampulla of Vater, it is necessary to establish large pooling projects to study their risk factors. This would be the largest prospective analysis to date for biliary tract cancers and could provide critical information toward establishing whether hypothesized risk factors are associated with risk and how those associations vary by site.

1.Investigate the association of BMI, smoking, diabetes, anti-inflammatory drugs, and family history of cancer with risk of biliary tract cancer.
2. Assess potential effect modifiers of these risks factors (e.g., sex, age, history of cholesterol gallstones).
3. Assess variation in risk by histological type (where appropriate)

Data on relevant exposures and covariates will be defined using standard categories and definitions across cohorts. BMI will be defined according to World Health Organization criteria: underweight, ≤ 18.49 kg/m2; normal weight, 18.5-24.9 kg/m2; overweight, 25-29.9 kg/m2; and obese, > 30 kg/m2 (14). Smoking will be evaluated by never/former/current smoking, as well as duration and intensity, where available. Diabetes will be treated as a yes/no variable. Ever use of anti-inflammatory drugs will be assessed, as well as duration, where available. We will also examine family history of breast cancer, prostate cancer, liver cancer, and all cancers. Relative risks and 95% confidence intervals for the association between these risk factors and cancer will be examined separately by sex and cohort using Cox proportional hazards models. If there is no significant heterogeneity between Cohorts, the data will be subsequently combined using meta-analysis techniques.

Few prospective studies have sufficient power to investigate the relationship between risk factors and biliary tract cancers in detail or within population subgroups. Pooled prospective studies such as those included in the LCPP would provide a unique opportunity to study these questions.

N/A

Primary, incident gallbladder cancer (C23.9) and other biliary tract tumors (C24.0, C24.1, C24.8, and C24.9)
o Date of diagnosis
o Histology: Include all. Adenocarcinomas will be distinguished from non-adenocarcinomas upon receipt of the data.
Method of cancer ascertainment (eg, cancer registry linkage, medical record, self-report)
Date of first diagnosis of any primary cancer (other than non-melanoma skin cancer)
Date of death, date lost-to-follow-up or study end date

Data previously requested/provided for LCPP
Baseline data on weight
Height
Waist/hip circumference
Smoking (status, duration, intensity, pack-years, age started smoking, age quit smoking, pack-years)
Diabetes
Anti-inflammatory drug use
Family history of cancer (name type of cancer)

Data previously requested/provided for LCPP
Date of birth
Age at baseline
Sex
Race/ethnicity
Education
Cigarette smoking
Alcohol intake
Physical activity
Previous history of cancer
New data requests (if available)
Previous history of benign thyroid diseases
Previous history gallstones
Previous history ulcerative colitis

No