The Power of Numbers: Melding Genomics and Epidemiology

NCI is moving on a number of fronts to harness the power of new genomic technology through epidemiologic studies designed to uncover gene variants that contribute to cancer susceptibility. Findings from NCI's portfolio of family studies have formed the basis for our understanding of many high-penetrant cancer-causing mutations. These rare mutations give unprecedented insights into carcinogenic mechanisms, but are responsible for only a small proportion of all cancer. Most cancer risk is believed to be due to gene-environment interactions involving low-penetrant but common genetic variants or single nucleotide polymorphisms (SNPs).

As outlined at last week's meeting of the National Cancer Advisory Board, some of the most promising of these studies of susceptibility genes are being led by investigators in NCI's Division of Cancer Epidemiology and Genetics (DCEG). Together with colleagues in the Division of Cancer Control and Population Sciences (DCCPS) and Office of Cancer Genomics, they have joined forces with extramural collaborators to form strategic partnerships that link epidemiologists, genomicists, and other investigators from the clinical, basic, and population sciences. This transdisciplinary team-based approach responds to a growing consensus in the scientific community that realizing the full potential of genomic and other new technologies requires their integration into robust, large-scale epidemiologic studies.

This can be accomplished through consortia that combine resources of several large cohort and/or case-control studies in a coordinated approach that allows rapid replication of positive findings. One such unique partnership is the Cohort Consortium, an international collaboration of intramural and extramural investigators responsible for 23 population cohorts encompassing 1.2 million individuals. The Cohort Consortium provides an integrative framework for nested case-control studies of specific cancers arising within the cohorts to systematically evaluate molecular and biochemical biomarkers of susceptibility and disease.

DCEG investigators are attempting to identify cancer susceptibility genes using the pathway-driven approach and genome-wide association studies (GWAS). The former approach involves studies that investigate risks associated with candidate genes selected on the basis of known function or biologic plausibility. For example, in a case-control study conducted in Spain and published last year in Lancet, DCEG's Dr. Montserrat Garcia-Closas and colleagues at NCI and in Spain showed that SNPs in two candidate genes, NAT2 and GSTM1, were associated with increased risk of bladder cancer and modified the risk due to cigarette smoking. Although the relative risks were modest, these SNPs are common among bladder cancer patients and may contribute to nearly one-third of the cases.

In contrast, GWAS use the latest technologies to "interrogate" the entire genome without targeting any specific gene. Led by a team of DCEG and extramural epidemiologists, statisticians, and genomicists, the Cancer Genetic Markers of Susceptibility (CGEMS) project is using high-throughput technology to identify and validate susceptibility genes for breast and prostate cancer. The potential associations identified will be validated in subsequent studies to eliminate false-positives due to chance. To date, GWAS have been completed on 1,200 prostate cancer cases and controls, and the data have been made available on NCI's cancer Biomedical Informatics Grid™ (caBIG™). It is anticipated that the initial scan for breast cancer will be posted in early 2007, and plans are underway to launch a similar scan for pancreatic cancer.

Dr. John E. Niederhuber
Director, National Cancer Institute