M4 - Whole genome scan for pancreatic cancer risk in the pancreatic cancer cohort consortium (PANSCAN)

Investigator Names and Contact Information

Rachael Stolzenberg-Solomon (rachael.solomon@nih.gov)

Introduction/Intent

Within the framework of the NCI-sponsored Cohort Consortium, a Pancreatic Cancer Cohort Consortium (PanScan) was formed to conduct a genome-wide association study (GWAS) of common genetic variants to identify genetic markers of susceptibility to pancreatic cancer. Specifically, a dense set of the most common genetic variants in the human genome, single nucleotide polymorphisms (SNPs) with sufficiently high enough minor allele frequencies (MAF > 5%) will be analyzed. The panel of SNPs is based on an analysis of common SNPs in individuals of northern European background determined by the International HapMap Project and provides an opportunity to monitor tested and untested SNPs because of linkage disequilibrium in the genome. It is estimated that the current panel of markers for a whole genome scan includes 550,000 SNPs and serves as markers for approximately 90% of all common SNPs in Caucasians.

In Phase 3, PanScan will conduct a new GWAS of ~1,600 recently identified incident pancreatic cancer cases with an equal number of controls drawn from 19 cohorts from the cohort consortium, including the 12 prospective cohorts who participated in PanScan 1 and 2, and seven additional cohorts. A joint analysis of the newly scanned cases will be conducted with cases from PanScan1 and 2 to identify novel regions of the genome associated with pancreatic cancer susceptibility. NCI will also conduct a GWAS analysis of pancreatic cancer survival using cases from both the cohorts and case-control studies. With the larger sample size (~5,500 cases and ~13,500 controls), it is anticipated that new genetic risk variants for etiology, and perhaps survival, will be identified.

Aims

1. To conduct a GWAS study of pancreatic cancer by genotyping approximately 1,600 recently identified incident pancreatic cancer cases and controls with previous GWAS data drawn from cohort studies of the NCI cohort consortium.
2. The genotype results and executive summaries of individual SNP analyses will be made available to bonafide researchers through a restricted-access website within several months of completion and following assessment of quality control.
3. To conduct a GWAS analysis of pancreatic cancer survival using all cases from the cohorts.
4. To conduct natural extension and value added studies of the PanScan GWAS which include:
   • a. As the SNPs identified from the GWAS will be markers for the causal variants, additional analysis for fine mapping, resequencing, and functional characterization of causal variant (s)
   • b. Candidate gene pathways analyses
   • c. Evaluation of gene-gene, gene-exposure and gene-biomarker interactions

Initially, in PanScan 1, 523,345 single nucleotide polymorphisms were genotyped in 1,528 incident cases and 1,594 controls from nested case-control studies of 12 cohorts plus 368 cases and 345 controls from one hospital-based case-control study. Taqman replication of SNPs from the three most promising regions in the initial scan, as determined by rank P values, was done in 2,457 cases and 2,654 controls from eight case-control studies that were part of the PanC4 consortium. A combined analysis identified an association between a locus on 9q34 and pancreatic cancer, the single nucleotide polymorphism rs505922 (P= 5.37 x 10-8; multiplicative per-allele odds ratio (OR) 1.20; 95% CI 1.12-1.28), which maps to the first intron of the ABO blood group gene whose protein determines an individual’s blood group (1). SNPs in the Sonic Hedgehog gene (SHH, chromosome 7q36.3) were promising in the cohorts (P values <10-7), but this signal did not replicate in the case-control studies.

Recently, the PanScan group conducted two GWAS that led to the discovery of four novel regions in the genome associated with risk for pancreatic adenocarcinoma. The first two scans were conducted in 12 cohort studies and 8 case-control studies. Since GWAS are scalable, particularly for the discovery of novel regions with moderate estimated effect sizes, the opportunity to increase the sample size will enhance the power of the study and therefore the probability of identifying additional novel regions of the genome associated with risk of this devastating disease.

Results/Findings

See Publications: 875, 930-934, 936, 1075, 1201, 1530, 1588, 1818. For a complete, up-to-date list of WHI papers related to this ancillary study, please use the searchable Papers section of this website.