APOLLO

Cancer research and care generate large volumes of complex data that could be used more effectively to diagnose the disease precisely, predict the patient’s prognosis and select the most effective therapy for a patient’s unique disease profile. To harness the power of this “big data,” a new environment is needed to promote cross-disciplinary collaboration among MD Anderson’s analytical labs, information systems, clinical care and research activities.

The APOLLO (Adaptive Patient-Oriented Longitudinal Learning and Optimization) program combines new and optimized existing workflows and infrastructure to create a more cohesive system for standardizing long-term collection of patients’ medical history and data derived from the patients’ biological samples. It’s coupled with research data and aggregated in a centralized big data warehouse for downstream analyses.

This adaptive learning environment relies on cutting-edge information and computer technologies to seamlessly blend patient data with knowledge from research studies plus best practices in clinical care to enable rapid and continuous learning – a process known as translation. APOLLO allows clinicians to benefit from the latest research insights, apply it against the complex patient data and clinical knowledge to generate better informed treatment decisions or hypotheses for testing. The goal is to improve the effectiveness of our care today and advance the future care for our patients.

Progress

To demonstrate the feasibility of this adaptive learning environment at MD Anderson, a pilot project in leukemia was launched in October 2012. To date, this pilot has accrued clinical data from more than 1,300 newly diagnosed leukemia patients. More than 3,000 patient samples have been collected using standardized procedures, and the resulting longitudinally structured and unstructured data are gathered in a centralized big data warehouse so advanced analytics can be applied to extract new discoveries and insights that can optimize treatments and improve outcomes.

In particular, one advanced analytical tool being developed is MD Anderson’s Oncology Expert Advisor™ (OEA), powered by IBM Watson. Watson is a third-generation cognitive computing system that won the Jeopardy! game show. IBM and MD Anderson are developing OEA as a cognitive clinical decision support tool that can provide oncology knowledge and expertise on demand tailored to a specific patient.

In the current pilot project on leukemia, OEA is exhibiting capabilities to: read and remember millions of pages of medical literature and practice guidelines; rank potential treatment options and care pathways based on the most up-to-date evidence; and learn through iterative training by MD Anderson clinical experts. The OEA solution for MDS-AML, a subtype of leukemia, has been launched for testing and evaluation, while it’s being expanded to support other major subtypes of leukemia.

Moving forward, MD Anderson intends to expand the APOLLO adaptive learning environment to include greater and greater amounts of big data and analytical tools like OEA to analyze those expanded data sets. The resulting insights will transform how we understand and treat cancer to reduce its burden on humanity around the world.

Lynda Chin, M.D.; Chair, Genomic Medicine; Research Engine
Lynda Chin, M.D. Chair, Genomic Medicine and Scientific
Director, Institute for Applied Cancer Science
Andrew Futreal, Ph.D., Moon Shots Program Co-leader; Professor, Genomic Medicine
Andrew Futreal, Ph.D. Program Co-leader
Professor, Genomic Medicine