AS681 - Metabolomic profiles and pancreatic cancer Risk in Women’s Health Initiative

Investigator Names and Contact Information

Rachael Solomon (rachael.solomon@nih.gov)

Introduction/Intent

Pancreatic cancer is a leading cause of cancer deaths, ranking third for mortality in the United States. Its incidence is increasing in the United States (1) and worldwide since the mid-1990s, particularly in developed countries (2). As there are no effective screening methods for detection of this malignancy, it is typically diagnosed at advanced stages, which contributes to a five-year survival rate of 9.3 percent (3). Most pancreatic cancers are ductal adenocarcinomas (PDAC)(4). The estimated timeframe from initiation of the pancreatic carcinogenic process to cancer death in humans is ~21 years (5). The progression between the first mutation to cancer is thought to be ~11 years, followed by most cases living with subclinical cancer for ~10 years prior to diagnosis (5). Smoking, a history of diabetes mellitus, adiposity, and chronic pancreatitis (6) are among the few established risk factors (7). Heavy alcohol use (8, 9), high total and red meat intake (10), high fat intake(11), a healthy diet (12) and a lifestyle index (13) have also been associated with PDAC. The diabetes associated with PDAC may have more heterogeneity than previously appreciated. Diabetes secondary to exocrine pancreatic disease (Type 3c) and common SNPs in genes related to Maturity Onset Diabetes of the Young (MODY) contribute to the association between diabetes and PDAC, in addition to Type 2 diabetes (14, 15). Genetic susceptibility influences PDAC, with familial pancreatic cancer estimated to account for 4-10% of cases, and an estimated heritability based on common susceptibility loci is up to 21% (16). Additional risk factors likely play a role; however, the evidence for specific exposures is often only suggestive due to the cancer’s low incidence and inherent methodological issues related to collecting data for this rapidly fatal gastrointestinal cancer. PDAC cases often have gastrointestinal problems, weight loss, or diabetes prior to being diagnosed due to sub-clinical cancer which can influence dietary intake and biomarkers. Prospective studies with prediagnostic measures of exposures are less prone to these issues.

The Specific Aims of this study are:

1. To identify metabolites associated with PDAC. We will conduct an association study, using a non-targeted approach, to replicate and identify new plasma metabolites (DiscoveryHD4 and CLP-lipidomics) associated with PDAC. Results will be meta-analyzed with the nested case-control metabolite studies from ATBC and PLCO that we have already measured metabolites. We hypothesize unique metabolites and metabolic profiles will be associated with PDAC.

2. To identify groups of metabolites, metabolic pathways, and lipid species metabolites associated with diabetes classification categories within WHI including subclinical PDAC, MODY genes, type 2 diabetes associated with obesity-insulin resistance and exposures (smoking and heavy alcohol use) known to cause diabetes related to pancreatogenic disease. We will use metabolite and lipid species data from WHI in aim 1 to conduct an untargeted analysis to identify metabolites associated with diabetes. We hypothesize that unique metabolites, metabolic pathways, and lipid species will be associated with each diabetes classification categories.

3. Examine the association between groups of metabolites identified in aim two and PDAC risk. Similar to aim 1, these analyses will be conducted within each cohort and meta-analyzed. We hypothesize that unique groups of metabolites, metabolic pathways, and lipid species that characterize different diabetes classifications will be associated with PDAC.

4. To identify metabolic pathways or biochemical/chemical groups associated with PDAC. We will evaluate the association between pathways and predefined biochemical/chemical class and PDAC risk. We hypothesize that metabolic pathways and biochemical/chemical groups will be associated with PDAC.

Secondary Aim: To identify groups of metabolites that connect known risk factors, such as overweight/obesity, insulin resistance, alcohol use, and smoking habits with PDAC and conduct mediation analyses of known risk factors. Previous studies have identified metabolomic profiles that characterize obesity, insulin resistance, diabetes, smoking, and dietary exposures. Using the metabolite data in Primary Aim 1, we will conduct a targeted analysis of metabolites that characterize exposures and risk of PDAC and conduct mediation analyses of known risk factors. We hypothesize that metabolomic profiles that characterize insulin resistance and other exposures will be associated with PDAC