Cancer is a family of over one hundred diseases. It is one form of abnormal growth of human cells. Benign and malignant tumors are both the result of abnormal growth, but only the latter is cancer and is capable of spreading throughout the body. Cancer risk is widespread; it affects 1 in 2 men and 1 in 3 women in their lifetimes. Susceptibility to the disease is a mixture of genetic, environmental, and behavioral factors.
One of the major goals of the Cancer Moonshot is to better understand cancer susceptibility factors and their interactions in order to develop tailored prevention and treatment programs for patients. The precipitous drop in the cost of sequencing the human genome has enabled this research and treatment. In 2003, the Human Genome Project spent $450 million to sequence the first complete human genome. By 2006 the cost fell to $20 – 25 million; in 2015, it cost less than $1,500. The Cancer Moonshot will take advantage of this new reality, as detailed below.
DoE and NCI partnership to accelerate precision oncology
Precision oncology is the idea of determining an individual patient’s DNA to tailor cancer therapies. The basic procedure can be outlined in three steps: characterize the genomes of a patient’s tumor; evaluate this data in light of knowledge of existing treatment methods and their fit for such genomes; and provide this evaluation to doctors to incorporate in a treatment program. Such customized treatments are predicted to become the norm in many areas of medicine, but it has been said that “oncology arguably sits at its vanguard” because “cancer is a genomic disease: most cancers harbor a cocktail of mutated … oncogenes [a gene that could turn a cell tumorous] and tumor suppressors.”
But a number of criticisms exist. As one scientist argued in Nature, “precision oncology has not been shown to work, and perhaps it never will.” More specifically, the author noted that there are not currently any targeted therapies offered, the genes found that seem to be relevant to the cancer are not influenced by treatment, and his literature review only uncovered thirty two instances of a response to such treatment by tumors over all the years this idea has been pursued.
The Journal of Clinical Oncology and The Lancet have both published comprehensive summaries of the state of this technology.
Applied Proteogenomics Organizational Learning and Outcomes consortium (APOLLO)
Proteogenomics looks at the expression of genes through proteins and is an intersection of proteomics and genomics. It is a new frontier of research, having only begun around 2008. Proteomics is the study of the proteins produced by an organism, just as genomics is the study of the genes of an organism. Proteogenomics is an improvement to the process of genome annotation, the identification of genes and determination of their function, because automated genome annotation can lead to erroneous results. Proteogenomics can improve that process, which in turn can provide novel ways to address cancer.
A technical discussion of the subject is available from New York University while an applied version is available in Nature Methods.
Big Data Analysis
Another front in the Cancer Moonshot is big data analysis of existing databases. Big data began as a field in the early 2000s due to technological developments. It has been applied to fields like business sales, fraud protection, and medicine. Big data, among other oncological applications, has been predicted to help identify the most effective cancer drugs to treat patients or even to recommend non-cancer drugs which may be effective treatments.
NCI Genomic Data Commons
A recurrent theme in the Cancer Moonshot is the application of existing stores of data for new purposes using big data analytics techniques. As an example, the GDC began by combining two petabytes of data from two existing NCI programs, The Cancer Genome Atlas and its pediatric equivalent Therapeutically Applicable Research to Generate Effective Treatments. A petabyte is equivalent to 13.3 years of HD video or 500 billion printed pages of text.
Within the expanded GDC, algorithms will standardize the data to allow easier use by researchers. The hope is that the data can be used to identify relevant genetic markers that will inform diagnostics and treatments.
DoD longitudinal precision oncology study
There are approximately 1,000 new cases of cancer in active duty military personnel per year and approximately 250,000 samples available in the DoD serum and cancer registries. These long-term samples provide a robust dataset for longitudinal oncology study.
As used in the longitudinal study, protein signatures are computational predictive models used to classify proteins into families. Such classification provides a basis for predicting which functional and structural properties may be found in a protein that has heretofore not been experimentally characterized.
In the epidemiological context, a longitudinal study is a cohort study documenting different reactions over time in a group which shares a common characteristic. For example, such a study could investigate differences in a collection of men, all of whom have smoked a pack of cigarettes each day of their lives and were born in 1975, but only some of whom develop lung cancer or emphysema, to determine which factors contribute to disease susceptibility.
The Million Veteran Initiative
This program is an aspect of the Precision Medicine Initiative (SᴄɪPᴏʟ brief available). The Initiative aims to move beyond the traditional “one-size-fits-all-approach” in medicine toward customized treatments tailored using an individual’s genes, environmental factors, and lifestyle.
Genes regulate the cycle of cell growth and death. Cancer occurs when a mutation, or several, converge to create uncontrolled cell growth. Importantly, similar mutations can cause cancers in different parts of the body, so a treatment focusing on the same set of mutations can sometimes be applied to multiple kinds of cancer. This search in tailored medicine for correlations between genetic mutations and disease outcomes has been criticized as suffering from the research problem of confusing correlation with causation.