Radiation oncologists dream of a day when, faced with a new patient sitting in their office, they can quickly consult a computer database offering specific treatment recommendations based on accurate, freshly updated data from millions of previously treated patients with cancer. To hasten that day, radiation oncologists and physicists speaking at the 2015 Big Data Workshop called for increased collaboration between medical societies, electronic medical record system vendors, and federal agencies to develop common standards and interoperable systems to enable databases to combine information.
“Radiation oncology is well positioned to harness big data to assess quality, facilitate research, and enhance patient care, because our treatment planning and delivery systems collect digital, structured information about our patients, and we routinely monitor our patients during treatment and in long-term follow-up,” said Karen E. Hoffman, MD, MHSc, MPH, Assistant Professor of Radiation Oncology, MD Anderson Cancer Center, and co-chair of the 2015 Big Data Workshop. Held at the National Institutes of Health (NIH), the workshop was sponsored by the American Society for Radiation Oncology (ASTRO), the National Cancer Institute (NCI), and the American Association of Physicists in Medicine (AAPM).
Radiation Oncology Registries
Despite their potential, existing registries and safety reporting systems specific to radiation oncology are few, limited, and incapable of exchanging data easily, workshop speakers noted. A closer look at several related projects discussed at the workshop follows:
The ASTRO-sponsored National Radiation Oncology Registry (NROR) recently completed a 1-year pilot project using prostate cancer as the initial disease site. Although 30 pilot centers were selected to participate, 11 pulled out due to legal or administrative concerns, and 5 sites did not participate because of data transfer issues, said Jason A. Efstathiou, MD, DPhil, Associate Professor and Director, Genitourinary Division, Department of Radiation Oncology, Massachusetts General Hospital, Boston. The 14 participating sites provided complete data for 430 patients, but it required significant manual data entry.
“ASTRO is continuing to monitor the field and considering the best options for future registry efforts, including collaboration with other specialty societies,” Dr. Efstathiou said. ASTRO is currently working with the American Association of Neurological Surgeons on a stereotactic radiosurgery registry and considering a possible collaboration with ASCO’s CancerLinQ™.
The Radiation Oncology Incident Learning System, launched in June 2014 by ASTRO and AAPM, is an online portal that allows radiation oncology centers to provide nonpatient-specific data about safety incidents at their facility in a secure, nonpunitive environment. More than 100 facilities are reporting safety events into this system, Dr. Hoffman said in an interview. “There is a great opportunity to integrate structured data collection into the clinical workflow and to generate comprehensive data, which improves the lifesaving radiation therapy we provide to more than one million cancer patients each year,” she said.
The Cancer Imaging Archive, a project funded by NCI’s Cancer Imaging Program, collects and curates radiation therapy information that can be combined across trials to produce larger research cohorts. The curation process improves data integrity and optimizes the potential for data reuse. The original goal of the project was to collect images from clinical trials to support research reproducibility, said principal investigator Fred W. Prior, PhD, Director, Electronic Radiology Laboratory, Mallinckrodt Institute of Radiology, Washington University, St. Louis.
Potential ASTRO-ASCO Collaboration
Ronald C. Chen, MD, MPH, Associate Professor of Radiation Oncology, University of North Carolina School of Medicine, Chapel Hill, who moderated a breakout session on the potential for big data in clinical decision-making and personalized medicine, said he favored the creation of a large cancer registry that would contain treatment details and quality measures for surgery, radiotherapy, and systemic therapy.
“ASTRO’s experience with NROR led to a recognition that each society developing its own registry was not an efficient use of resources. Collaboration makes much more sense,” said Dr. Chen.
A big data system for radiation oncology should capture the anatomic area treated, the radiation technique used, and the dose delivered, Dr. Chen added. Also, ASTRO’s guidelines and quality measures could be included in a comprehensive oncology registry, he said.
ASCO and ASTRO have discussed collaboration, according to Robert S. Miller, MD, FACP, Senior Director, Quality and Guidelines, ASCO, and Medical Director, CancerLinQ (see sidebar). “We are certainly very open to that possibility,” Dr. Miller said. “ASTRO is a very logical partner for ASCO in this effort.”
CancerLinQ has begun working with the first of 15 oncology practices selected for the system’s beta launch. Some of the practices include radiation oncology components, said Dr. Miller, who spoke at the workshop about CancerLinQ. “We are not necessarily set up from an informatics standpoint to map detailed radiation oncology treatment planning information just yet, but there’s nothing that would stop us from doing that,” he explained.
“Most of our early focus with CancerLinQ are aspects of clinical care that are applicable to either specialty,” Dr. Miller added. “The system is not tuned to one particular mode of practice or specialty. We can bring in anything from anywhere. It’s all about getting as many charts as possible into CancerLinQ. The more data you aggregate, the better.”
ASTRO is planning a a Precision Medicine Workshop scheduled for June 16–17, 2016, on the NIH campus in Bethesda, Maryland. ■
Disclosure: Drs. Hoffman, Efstathiou, Chen, Prior, and Miller reported no potential conflicts of interest.
