Prescription for Progress in Cancer Prevention
By Peter G. Shields, MD,
deputy director, The Ohio State University Comprehensive Cancer Center –
Arthur G. James Cancer Hospital and Richard J. Solove Research Institute, member of the OSUCCC – James Cancer Control Programdirector of the Prostate and Genitourinary Oncology Clinic, leader of the OSUCCC – James Molecular Carcinogenesis and Chemoprevention Program, and a member of the Committee to Review Dietary Reference Intakes for Vitamin D and Calcium
Although medical science has made great strides in many areas of cancer research and treatment, progress in cancer prevention has been far too slow and somewhat disappointing. Simply put, far too many people still get cancer. We know a lot about the causes of cancer, such as diet, sunlight, tobacco, alcohol and physical inactivity, but we have known about most risk factors for decades. At the same time, new epidemiology studies make only incremental advances, while subsequent intervention studies often do not validate the expected outcomes. Consequently, early detection remains the best, and often the only, option for reducing the cancer burden. This is not good enough.
For real progress in cancer prevention, we need a deeper understanding of cancer's causes, then we must use that understanding to develop clinically useful markers of cancer risk and evidence-based reasons for choosing a particular intervention or combination of interventions. As we well know, cancer develops from a number of natural cell mechanisms that have gone awry. We need a better understanding of the pivotal nodal points that cause the cascading deregulation of these multiple mechanisms. These nodal points then offer logical targets for intervention.
We can meet these needs using the latest technologies and a three-part prescription for a broader research strategy. It begins with a systems-biology approach that involves multiple 'omics assessments, e.g., genomics, epigenomics, transcriptomics, proteomics and metabolomics within a single study, and it is guided by experimental studies that follow the continuum from normal cellular function to cancer.
The second part is to take the multiple 'omics data and vertically integrate them within the experiment studies to see what pathways are affected differently but combine to move the cell toward cancer. These findings are then corroborated across different study designs, using a variety of human cell-culture models, epidemiological studies of healthy subjects and cancer patients, and clinical trials. Thus, the added component in this second part is horizontal integration of the data.
Critical to this process is the identification of how widely people vary (interindividual variation), which provides clues by considering those individuals we usually call the outliers. With this approach, we identify the most promising biomarkers of cancer risk and targets for prevention.
In 2010, more than 1.5 million new cases of cancer were diagnosed and 569,490 persons died of cancer in the United States.
• This amounts to more than 1,500 Americans dying each day and makes cancer the second leading cause of death, after heart disease.
• African-Americans are about 33 percent more likely to die of cancer than are whites and more than two times more likely to die of cancer than are Asian or Pacific Islanders, American Indians and Hispanics.
• According to the National Cancer Institute, the estimated total cost of cancer in 2005 in the United States was $209.9 billion.
The third part is to validate the importance of these biomarkers by first assessing them in molecular epidemiology studies and then in cohort and tailored intervention studies. In the future, we should have validated clinical markers that improve predictive capabilities and shift the typical disease timeline toward earlier interventions
that will be more effective and will minimize the need for later treatment.
Taking this "nodal" approach should produce findings of high statistical significance, reduce the number of false positive findings, assist in separating out "noise" and downstream effects inherent in 'omics analysis, and reveal how biological mechanisms interact to make normal cellular mechanisms go awry.
Breast cancer exemplifies both the problem and how this approach might work. Breast cancer is the most common malignancy in women. Incidence rates have risen slowly for the past two decades, and the disease remains the second leading cause of cancer-related death in women.
Of the 209,000 women diagnosed yearly, 80 to 90 percent are sporadic cases (i.e., women without a strong family history of breast cancer). For these women, known risk factors such as age, reproductive or hormonal risk factors, hormone replacement therapy and alcohol consumption account for less than 50 percent of the risk. Given this, we have insufficient methods to predict which individuals in the general population are likely to develop the disease, and therefore we do not do a sufficient job at personalizing cancer prevention.
To prevent breast cancer, we need molecular markers that enable us to identify women at high risk. These women can then work with their physicians to take a proactive, personalized preventive approach, such as behavioral interventions for lifestyle and chemoprevention. They also can have tailored screening methods to shift detection to earlier time points when treatments are most effective.