CER “Cyberinfrastructure” May Boost Cancer Research

June 2010, Vol 1, No 2 - Comparative Effectiveness

Researchers from the University of California and the University of California San Diego (UCSD) have begun working on a prototype “cyberinfrastructure” that would allow cancer researchers to collect and interpret data from multiple sources to compare the effectiveness of preventive measures, drugs, treatments, and interventions during the course of clinical trials.

The project, known as CYCORE (Cyberinfrastructure for Comparative Effectiveness Research), will aggregate data from clinical trials, patient medical records, self-reported and objectively monitored social and behavioral data, data on cancer outcomes from regional cancer registries, and cost-benefit analyses. The goal is to allow scientists to better collate ever-increasing amounts of data, and to use that both to impact the design of future trials and to positively affect patient outcomes in the present.

Collecting Real-world Data

A key part of the project will involve home and mobile phone-based sensing devices that will allow real-time capture of pertinent patient data. Kevin Patrick, MD, a professor in UCSD’s Department of Family and Preventive Medicine and a principal investigator in the project, emphasized the novelty of the approach in a press release discussing the effort. “A primary focus will be on tackling the problem of obtaining, in patients’ homes, objective person-level data on behaviors such as adherence to cancer medications, diet, physical activity, sleep, environmental exposures, and quality of life,” he said. These factors are wide ranging, encompassing everything from physiological matters (eg, cardiac function, blood pressure, sleep-related EEG data, etc), to behavior-related issues (eg, do they take their medication as prescribed, do they exercise, do they seem lethargic, are they still smoking, do they eat appropriate foods, etc), to environmental factors (eg, is there secondhand smoke in the home, is their neighborhood experiencing significant amounts of air pollution, is the patient always alone, etc).

“These factors are enormously important in the course of medical treatment, but are almost always assessed through infrequent and after-the-fact self-reports,” Dr Patrick added. “To have these data in near-continuous form will greatly expand our understanding of who does well on what treatment, when, why, and for how long.”

Laura Wolszon, PhD, a program director on the project, elaborated, saying “one of the reasons that cancer treatments haven’t progressed further is because researchers don’t typically evaluate the impact of a patient’s behavior and personal environment when deciding whether one treatment works better than another. The more data we use, the more accurate and effective will be the treatment plan for that patient and, probably, others.”

The researchers are constructing the prototype using data from their partner, M. D. Anderson Cancer Center, but the final goal is to include data from all cancer clinical trials. The research team emphasizes that the program will be open source, allowing others to modify it for their own purposes, and will be adaptive to changes in policy along the way—say, if changes in privacy laws occur that suddenly restrict access to certain data.