Honing the tools to identify clues to cancer survival

By Cindy Fox Aisen

McDonald

Drawing upon their long-established expertise in medical informatics, researchers from the IU School of Medicine will develop, organize and test secure databases with unidentifiable data which in the future will allow scientists to review, compare and contrast the pathology reports and the DNA and protein content of stored tissues from all Indianapolis hospitals.

Researchers from the Indiana University School of Medicine (IUSM) and the Regenstrief Institute for Health Care are developing the tools that may bring scientists closer to identifying that key through a grant of $7.4 million from the National Cancer Institute.

Drawing upon their long-established expertise in medical informatics, they will develop, organize and test secure databases with unidentifiable data which in the future will allow scientists to review, compare and contrast the pathology reports and the DNA and protein content of stored tissues from all Indianapolis hospitals.

Building this “machinery” and organizing the data is a challenging task that involves developing a natural language process, eliminating duplications, instituting standardization, and “scrubbing” the data to eliminate person and place identifiers, according to Dr. Clement McDonald, the study’s principal investigator and a internationally known medical informatics pioneer.

When the five-year project is completed, McDonald hopes to have a tool that will allow other researchers to compare DNA, proteins and other biological factors to determine differences between the same tissue with and without cancer, those tissues with primary cancers and metastasis, and the differing effects of therapy. McDonald is an IU Distinguished Professor and Regenstrief Professor of health services research at IUSM and director of the Regenstrief Institute for Health Care.

The new study will draw upon data from as many as 1.6 million pathology reports and 6 million tissue blocks from IU Hospital, Riley Hospital for Children, Methodist Hospital (which together comprise Clarian Health), Community Hospitals, St. Francis Hospitals, St. Vincent Hospitals and Wishard Memorial Hospital--all in Indianapolis. These institutions, which currently participate in a unique city-wide emergency room care collaboration with the Regenstrief Institute, account for 95 percent of the city’s hospitalizations and reported cancer cases.

This ability to access data from a whole population, which the IU and Regenstrief work will make possible, will provide future researchers with a unique opportunity to measure the importance of various factors in a large group exposed to the same or similar environmental factors.

“Long-term cancer survivors have something in their biology that is different from those who don’t survive. If we can provide the critical tools that will assist in determining what that is, we may enable scientists in the not-too-distant future to figure out how to alter the biology of the cancer patient, perhaps by activating or blocking a protein or its receptor, to save lives,” said McDonald. “This work may ultimately lead to studies resulting in the discovery of drugs that activate proteins or block receptor sites for biologic pathways,” he noted.

McDonald also heads the medical informatics program of the Indiana Genomics (INGEN) Initiative at IU. The fundamental research goal of INGEN is to establish the relationships between human genotypes and the physical and biological characteristics of tissues and organ systems.

The new Shared Pathology Informatics Network (SPIN) builds upon 25 years of experience with the internationally respected Regenstrief Medical Records System. RMRS is a database with more than 300 million laboratory results, radiology and pathology reports, diagnostic studies, operative notes and discharge summaries. SPIN will also draw upon expertise gained from Indianapolis Network for Patient Care (INPC), a data repository which stores encounter records, and clinical laboratory data for use in care at emergency rooms city-wide.

In his role as an investigator for the SPIN, McDonald will head a consortium of institutions including the Indianapolis hospitals, the Indiana State Department of Health and the University of Pittsburgh. The University of Pittsburgh group, known for its expertise in pathology informatics, is led by Dr. Michael Becich.

McDonald also will work with other investigators from hospitals affiliated with Harvard University and with the University of California at Los Angeles. Together, these institutions will develop a mechanism to share their medical data, while strictly preserving patient confidentiality. This will give many researchers their first opportunity to conduct large-scale database searches on medical data, asking basic questions that cannot be answered with small numbers of patients. Having access to large medical data sets will give researchers the best opportunity to make new cancer discoveries.