The Cancer Moonshot, which aims to cure cancer once and for all, needs to account for environmental factors — not just clinical ones — in order to succeed.
What will it take to “cure” cancer once and for all? The answer might be closer than we thought.
In December 2016, Congress authorized $1.8 billion in funding for Cancer Moonshot, an ambitious initiative established by the Obama Administration and led by the National Cancer Institute (NCI), a division of the Nation Institutes of Health. What’s more, NCI recently established a centralized bioinformatics Genomics Data Commons and advocated for the development of clinical bioinformatics in the NCI Annual Plan & Budget Proposal for Fiscal Year 2017. These investments brought renewed attention to a disease — or rather, a collection of related diseases — that afflicts about 40 percent of U.S. men and women at some point in their lives and kills 500,000 Americans each year.
The announcements have prompted discussion about how we can harness the massive amounts of data on molecular, genetic, immunologic, and other clinical aspects of cancer to improve diagnosis and treatment. These data are typically located in disparate repositories (clinical research studies, electronic health records (EHR)/personal health records (PHR), genomic sequencing labs), which means that data science and relevant analytic methods will be crucial in merging them and harvesting insights.
To be truly effective in confronting cancer, though, we need to focus more attention and resources on the behavioral and environmental aspects of cancer —things like smoking, preventive health, and quality of care. A recommendation for more robust behavioral informatics — which uses data to better understand human behavior and communication — was noticeably absent from the Genome Commons and other initiatives that support the ultimate Cancer Moonshot goal. The recent announcement by NCI and NIH to develop a Blue Ribbon Panel of scientific experts, cancer leaders, and patient advocates — designed to inform the Cancer Moonshot Initiative — does reference cancer prevention and includes public health scientists, such as Dr. Barbara Rimer, on the panel. However, the key themes of the initiative, though, mentions little about the importance of behavioral and environmental factors.
The Role of Behavioral and Environmental Factors in Cancer Research
While big data related to bioinformatics (e.g., genetic data) and clinical informatics (e.g., EHR/PHR data) are critical, behavioral and environmental big data related to this deadly disease are just as important. Some examples:
- Tobacco use, including e-cigarettes and smokeless tobacco
- Sun safety behaviors and exposure to ultraviolet radiation
- Obesity-related behaviors, like diet and physical inactivity
- Preventive health measures, such as vaccines and screenings
Behavioral and environmental factors like these affect individuals across the cancer continuum, from prevention to survivorship. Moreover, identifying phenotypic aspects of cancer involve assessing BOTH genetic and environmental (including behavior) factors. Prominent researchers and clinicians alike have lobbied for the importance of understanding genes and environment in tandem as we examine the cancer continuum (i.e., prevention, detection, treatment, survivorship).
Take cigarette smoking, for example – the leading cause of cancer and death from cancer according to the NCI. Almost 17 percent of U.S. adults currently smoke, with much higher rates among less educated and lower income groups (25+ percent), and particular ethnic minority populations like Native Americans (29 percent). Helpful electronic resources, like Smokefree.gov, are now available for those who are looking for information and tools designed to help them quit. For those diagnosed with smoking-related cancers, other resources are available. But impediments to progress — just 5 percent of eligible cancer patients participate in clinical trials and rates of tobacco smoking are predicted to increase internationally — have made it more important than ever to understand the behavioral factors at play.
And that’s just tobacco.
The behavior of patients and providers, both before and after a diagnosis, are relevant and critical, too. Did the provider inquire about a family history of cancer? Were healthcare professionals involved in a patient’s treatment plan dismissive or insensitive? Has the patient adhered to medication regimens? All of these behaviors — and many more — can impact health outcomes and quality of care, which means that we need to dedicate more resources to understanding how and why they manifest.
The Power of Better Big Data
We have the ability to collect more data about these factors than ever before. Federal support on this front has been encouraging — data science and big data both feature prominently in the NIH-Wide Strategic Plan Fiscal Years 2015-2020 — and technology is better poised than ever to collect massive amounts of data on these key behaviors and environmental exposures, sometimes in real-time and with the promise of measuring behavior and environments across large segments of the population. Patients are taking advantage, too: surveys have indicated that more patients are accessing information in their EHRs/PHRs, potentially empowering them with information to take control of their health.
In addition to collecting more data about behavioral and environmental factors, though, we need to make sure we can interpret it. We’ll need to develop more intuitive, user-friendly technology and data systems that enable cancer patients to access relevant health/medical information, increase patient-provider communication/interactions, and optimize personalized patient-centered care among multiple healthcare providers.
When it comes to methodology, the research community plays an important role, too. These communities must agree on measurement and data standards for behavioral and environmental assessments in these seminal health research initiatives. Future major research endeavors should integrate concepts of behavior change into various longitudinal studies, especially intervention research. Furthermore, figuring out the best ways to integrate genetic/biologic and behavioral/environment data from multiple uber-large datasets will be a critical challenge for the research community — meeting it will require collaboration and coordination across the gamut of stakeholders, including government agencies, nonprofit funding organizations, healthcare provider organizations, health information technology companies, and academic institutions — to name a few.
We’ve made some important strides, but there’s still ample work ahead. As the National Institutes of Health ramps up the Precision Medicine Initiative (PMI), with a large initial focus on cancer research, NIH and the various stakeholders in PMI need to curate informatics data on behavioral and environmental factors, just as much as genetic, molecular, and clinical factors. The success of our Cancer Moonshot — and the health of our population — depends on it.