The past decade has witnessed revolutionary breakthroughs in biology that will transform the treatment of cancer. The discovery of genes implicated in cancer and how patients respond to therapies already has led to targeted treatments for colorectal, breast and other cancers and has made the promise of using a patient’s characteristics to guide treatment over the course of his or her disease a central focus of the nation’s cancer research agenda.  

The underlying issue

As emphasized by the White House Precision Medicine Initiative, the key to achieving personalized cancer medicine is analyzing the huge amounts data that are being generated.

Vast amounts of data from individual patients collected in clinical trials and other health studies and from electronic health records (EHRs), patients themselves, and their providers, describing not only their genetic makeup but their demographics, physiology, lifestyle, environment and medical history must be integrated, analyzed and translated into tailored treatment strategies. But as noted by Dr. Francis Collins, director of the National Institutes of Health, in a 2013 Washington Post opinion article, “this mountain of data will be of limited use to cancer patients if researchers and clinicians lack the tools necessary to manipulate and mine it effectively.”

"Statistical scientists are developing innovative new techniques to address the myriad challenges of using data to personalize treatment."

Factoring in

Statistics, the science of learning from data, is providing these essential tools. Statistical methods for the design and analysis of cancer clinical trials have been fundamental to cancer research for decades. Today, statistical scientists are developing innovative new techniques to address the myriad challenges of using data to personalize treatment.  

New ways to design clinical trials that incorporate biomarkers—characteristics reflecting a patient’s disease status—will enable the identification of subgroups of patients to whom a certain therapy can be targeted. Statistical methods are being developed for synthesis of data from disparate sources to determine which studies to conduct next, and in which patients, and for converting complex data from EHRs into knowledge.

Moving targets

New statistical approaches for harnessing the breathtaking volume of genomic and other data, generated from continually evolving technologies, into insight on potential biomarkers and candidate individualized cancer therapies are also being created.

Additionally, statistical techniques are being developed to turn data into personalized strategies for guiding the therapeutic decisions that must be made at milestones in the progression of a patient’s disease. These strides allow experts to optimize a patient’s long-term health outcome while minimizing the chance of serious side effects, as well as clinical trial designs to provide data specific to this purpose, and will provide cancer clinicians with evidence-based decision support.

Statistical scientists are working in close collaboration with clinical, biological and computational scientists to conceive and create these and other methods. Through this work, statisticians are playing an integral role in making personalized cancer medicine a reality.