"Toward the Last Cohort"
Dr. Potter presented a proposal for a large cohort study that aims, among other things, to characterize the genetic variation and environmental exposures involved in different types of cancer and to identify biological markers for the early detection of disease. Another goal is to expand the molecular classification of tumors.
To achieve all this, Dr. Potter said he would need a very large cohort of ethnically diverse individuals who are well characterized genetically, whose exposures are diverse and carefully documented, and whose illnesses and mortality can be monitored.
"A hundred thousand people would not be enough," said Dr. Potter. "A million might be a good start."
This group, "the last cohort," could be used to examine the impact of exposures on the causes and rates of disease and also to study the interaction of these exposures with genetic variation. The project, which would focus on cancer but could be a model for other diseases, would define a "pattern of human disease susceptibility and resistance."
The project would collect biological samples from individuals - to define DNA sequences and protein expression - and information about lifestyle and environmental exposures, including everything from medical history, family history, and reproductive history to data on diet, smoking, hormone use, and exercise.
The exposure data are key, Dr. Potter emphasized. He said that there will be remarkable high-throughput gene sequencing and proteomics capabilities in the next 5 or 10 years, but there is also a risk for not having the infrastructure in place to collect the data on exposures, thereby compromising our ability to use the combination of information to understand and treat disease.
One goal of the cohort study would be to continue the recent progress in developing a system for the molecular classification of tumor subtypes and creating more precise definitions of disease, or phenotypes. In general, the current system for characterizing disease runs from precise molecular classifications of cancer to vague heterogeneous syndromes such as schizophrenia, Dr. Potter said.
"The disease classification system based on histology that has served us well since the 19th century is beginning to fall apart because it isn't helping us understand diseases and how to treat them," he said. "The good news," he added, "is that you can always improve the classification system as new information becomes available."
Dr. Potter said that the study offers the potential for discovering biological markers for early detection because individuals would be followed for a prolonged period of time and would donate blood and other bodily fluids at regular intervals. By the time patients are typically diagnosed and recruited for studies, most indicators of early-stage disease have come and gone.
Dr. Potter also explained that the variation in cancer rates among populations around the world cannot be accounted for by variation in genes alone. Rather, it must be due to differences in environmental exposures or to differences in the interactions between genes and exposures.
See also: "Toward the Last Cohort," an editorial in the June 2004 Cancer Epidemiology Biomarkers & Prevention.