Reflections on the Past and Future of Epidemiology

In May, DCEG was privileged to hold a panel discussion titled Cancer Epidemiology over the Last Half-century and Thoughts on the Future. The panel featured Joseph F. Fraumeni, Jr., M.D., and Dr. David Schottenfeld, University of Michigan, co-editors of multiple editions of the indispensable textbook Cancer Epidemiology and Prevention. Robert N. Hoover, M.D., Sc.D., Director of DCEG’s Epidemiology and Biostatistics Program, moderated the discussion and posed questions on the major influences that affected the careers of Drs. Fraumeni and Schottenfeld, the key contributors and seminal discoveries in the field of cancer epidemiology, and advice for young epidemiologists. The full-capacity audience of NCI staff thoroughly enjoyed the panel members’ thoughtful, gracious, enlightening, and often humorous responses.

Both Drs. Fraumeni and Schottenfeld initially had trained as clinicians, but influential mentors steered them toward careers in epidemiology. Dr. Schottenfeld was in his second year in the Epidemic Intelligence Service at the Centers for Disease Control and Prevention when he was assigned to work with Dr. Abraham Lilienfeld at the Johns Hopkins School of Hygiene and Public Health on a multi-hospital case-control study of male breast cancer. Dr. Schottenfeld recalled, “When I went to Hopkins, the chemistry with Abe Lilienfeld was perfect. On completing the project, I was persuaded to abandon my notion about being a practicing oncologist but instead pursue a career in cancer epidemiology and public health. The relationship that we developed was sustained throughout my career.”

Similarly, Dr. Fraumeni was encouraged to go into epidemiology by Dr. Rulon Rawson, chairman of the Department of Medicine at the Memorial Sloan- Kettering Cancer Center, where Dr. Fraumeni served as chief resident.

During weekly medical rounds, he was responsible for guiding Dr. Rawson to the most interesting cases on the wards. Dr. Fraumeni’s choices, however, led Dr. Rawson to remark, “You are thinking like an epidemiologist. It’s the patterns that you’re focusing on, and that’s epidemiology.” Dr. Rawson encouraged Dr. Fraumeni to consider spending time at NCI, where he was mentored by Dr. Robert W. Miller, then Chief of NCI’s Epidemiology Branch, who demonstrated the ways in which someone with a strong clinical orientation might contribute to epidemiologic research.

When asked to identify the cancer epidemiologists who have had the most significant impact on the field, Drs. Fraumeni and Schottenfeld agreed that the towering figure in cancer epidemiology, and epidemiology in general, was Sir Richard Doll at Oxford University in the United Kingdom. Professor Doll, along with Dr. Ernst Wynder in the United States, was the first to establish a causal relationship between cigarette smoking and lung cancer. However, Professor Doll’s contributions to epidemiology extended well beyond that landmark finding.

Dr. Fraumeni also cited Mr. William Haenszel, former Chief of NCI’s Biometry Branch, whose landmark studies of the changing cancer risks among migrant populations offered many key epidemiologic insights; Dr. Miller for his trailblazing work in clinical epidemiology and childhood cancer; and Dr. Alfred Knudson for his seminal contributions to cancer genetics. Dr. Schottenfeld also gave tribute to Sir Austin Bradford Hill. “His book Principles of Medical Statistics had an enormous impact on the medical profession, which tended not to think numerically or quantitatively,” Dr. Schottenfeld explained. He also credited other biostatisticians who have made important and lasting contributions by developing rigorous study methods and analytical approaches in epidemiology. They included Mr. Nathan Mantel, Mr. Jerome Cornfield, Dr. Norman Breslow, Mr. Nicholas Day, Dr. Kenneth Rothman, and Dr. Sander Greenland.

After describing the dominant influence of tobacco in cancer etiology, Dr. Fraumeni spoke about the changing themes that motivated epidemiologic research over time, such as the possible role of viral infections during the 1960s; environmental and occupational hazards during the 1970s; dietary practices, including fat intake, in the 1980s; and genetics from the 1990s on. For epidemiology, the genomics era began with family-based studies of rare, high-penetrant genes and vastly expanded as new technologies enabled genome-wide association studies to identify common low-penetrant genetic variants at the population level. Dr. Schottenfeld cited several remarkable discoveries in epidemiology that were transformative in nature, such as the roles of human papillomaviruses in cervical cancer, hepatitis B and C in liver cancer, asbestos exposure in lung cancer and mesothelioma, ionizing radiation in a wide variety of cancers, and the detection of low-level risks of lung cancer associated with passive smoking.

Drs. Fraumeni and Schottenfeld also discussed epidemiologic discoveries that were surprising at the time and caused paradigm shifts in thinking about cancer etiology. Among their examples were the association of Helicobacter pylori infection with duodenal ulcer and gastric cancer; the complex and varied effects of steroid hormones on the risk of breast and other cancers; the protective effects of aspirin on colon cancer risk; the discovery of vaginal adenocarcinomas in the daughters of women treated with diethylstilbestrol during pregnancy, which called attention to the potential importance of early-life exposures for cancer etiology; and the growing evidence that obesity increases the risk of several forms of cancer.

When asked about notable “characters” in cancer epidemiology and what made them so interesting, Drs. Fraumeni and Schottenfeld agreed that Dr. Wynder was the most colorful. Dr. Schottenfeld described him as “dramatic, flamboyant, charismatic, and, at times, infuriating. He had endless ideas. The only issue was that you had to pursue them. But he was enormously productive with hundreds of important publications.” Dr. Fraumeni remarked that Dr. Wynder combined a very active professional and social life, recalling a dramatic moment at the Memorial Sloan-Kettering Cancer Center when Dr. Wynder strolled into the cafeteria with the movie star Kim Novak.

After commenting on non-epidemiologists who had a positive effect on the field and discussing what they considered to be their own personal successes, Drs. Fraumeni and Schottenfeld turned the discussion to the importance of training the next generation of scientists, a career-long passion for both men. Dr. Schottenfeld exhorted young investigators, “You are going to have to be life-long students. You must continue to learn and continue to master your methodologies. You need to experience passion and joy in the research or else don’t trouble yourselves with all the years of competing for funding. You need to be questioning, to be critical observers, to think ‘out of the box’ in imaginative ways about questions, because it is framing the questions that’s the important issue.” Dr. Fraumeni encouraged young investigators to find the best possible mentors, work with compatible collaborators who have complementary skills, and carve out a niche in which to grow and eventually become an authority on the subject.

Drs. Fraumeni and Schottenfeld agreed that with imagination and resources, investigators should seize the opportunities to discover and investigate “natural experiments,” especially in the form of high-risk populations that provide etiologic clues. They also encouraged participation in large-scale collaborations and consortia when team science is warranted. Dr. Fraumeni concluded, “Despite the fact that we know so much more than we did 50 years ago, there is still an enormous amount that remains to be learned, and the opportunities in cancer epidemiology are greater than ever.”