by Patricia Hartge, Sc.D. and Wendy Schneider-Levinson
Collaboration is the hallmark of contemporary scientific research, as scientists form teams and partnerships to study complex, multifaceted problems. In epidemiology, the relatively small studies of 300 or fewer subjects that predominated before the mid-1980s would not have adequate statistical power to investigate the relatively low-level effects of risk factors that are of particular interest today. Under the leadership of Joseph F. Fraumeni, Jr., M.D., DCEG has partnered with NCI’s Division of Cancer Control and Population Sciences (DCCPS) to foster and support large-scale extramural-intramural consortial initiatives to assemble the substantial quantity of data and biospecimens necessary to conduct a wide range of epidemiologic studies.
An early example of “big science” involving extramural-intramural collaboration was the National Bladder Cancer Study, which DCEG launched in 1978 to examine the role of artificial sweeteners, notably saccharin, in the development of bladder cancer, a relationship suspected on the basis of laboratory animal studies. Because the expected risk was low, the investigators determined the need to enroll approximately 3,000 cases and 5,800 controls, about 10 times the size of a typical study of the era. Toward this end, DCEG investigators partnered with the Surveillance, Epidemiology and End Results (SEER) program to conduct a multicenter, population-based study. In addition to shedding light on the major public health concern about saccharin use, the study provided an opportunity to investigate the effects of a wider range of lifestyle, environmental, and occupational exposures on bladder cancer risk.
In recent decades, epidemiologic research has become increasingly interdisciplinary, especially with the advent of new genomic and molecular markers that have sharpened our measures of causal factors and mechanisms as well as diverse outcomes. Biomarkers of cancer susceptibility initially took the form of studies that evaluated the risk associated with candidate genes, which were underpowered and led to many positive findings that could not be reproduced. DCEG was well positioned to foster and participate in large-scale consortia involving case-control, cohort, and family studies, a strategy through which genetic findings could be identified with some precision and then replicated rapidly and efficiently.
In 2002, extramural and intramural leaders of case-control studies of non-Hodgkin lymphoma (NHL) formed the international “InterLymph Consortium” to investigate a number of candidate genes suspected to confer susceptibility (see Figure 1). The Consortium identified the TNF gene as playing a role in NHL and, just as importantly, it ruled out the IL1A gene as a meaningful contributor to risk. InterLymph has now grown to include consortia for genome-wide and other studies of Hodgkin lymphoma and multiple myeloma.
In 2003, the Breast and Prostate Cancer and Hormone-Related Gene Variant Study (BPC3) was launched to enable large-scale analyses of candidate genes that affect hormone metabolism. The NCI study combined the resources of two intramural cohorts and six extramural cohorts, with prospectively gathered prediagnostic plasma samples; lymphocytes; body measurements; and extensive questionnaire data on diet, physical activity, exogenous hormones, smoking, and other lifestyle factors from more than 740,000 men and women, including 8,850 with prostate cancer and 6,160 with breast cancer. With the availability of high-throughput genomic and metabolic technologies for genome-wide association studies, the epidemiologic infrastructure has helped to unravel the role of genetic and hormonal risk factors for these tumors.
Family studies–based consortia also have been a focus for research on inherited mutations and cancer susceptibility. One of the earliest examples is the GenoMel Consortium, formed in 1997, in which DCEG joined forces with groups from more than 20 countries. The collaboration aims to identify and understand the genetic underpinnings of melanoma, including interactions with sun exposure and other risk factors, and their incorporation in risk-prediction models for individual patients.
Consortial strategies have more than proved the worth of large-scale collaborative research that incorporates innovative high-throughput technologies into robust epidemiologic designs to dissect the genetic and environmental components underpinning complex disorders such as cancer. NCI currently supports more than 50 consortia of various kinds. “Dr. Fraumeni recognized early on that for the kind of detective work it takes to unravel the complexities of cancer etiology, you need two things—teams of experts (because no one person can have all the expertise) and access to large numbers of study subjects and measurements of data so you can study the multiplicity of potential risk factors,” said Deborah M. Winn, Ph.D., Deputy Director of DCCPS. To date, the combined efforts of DCEG and DCCPS have led to dozens of epidemiologic consortia that are striving to inform clinical and public health interventions.
When asked to evaluate the future of collaborative studies in the era of molecular epidemiology, Dr. Fraumeni commented, “There are many challenges in establishing large-scale interdisciplinary studies, but we need such strategies if we hope to more fully understand the biological mechanisms of cancer induction and progression and the means of preventing and controlling cancer.”