Health Information Technology Considerations for Oncologists, Genomic Tumor Boards


James Lin Chen, MD, Ohio State University, and chair, ASCO CancerLinQ Oncology Informatics Task Force, discussed the various challenges and considerations of enabling precision medicine with health information technology.

Dr Chen began his presentation by alluding to the fact that while the treatment paradigms have traditionally remained constant, clinical pathways are becoming increasingly complex. To illustrate this relationship, a comparison of NCCN practice guidelines for non-small cell lung cancer (NSCLC) in 2010 and 2018 were highlighted. In the 2010 edition, genomic profiling consisted of one biomarker (EGFR) and two medications. In contrast, the current 2018 edition has five biomarkers (EGFR, ALK, ROS1, BRAF, and PD-L1) and 13 therapies.

Meanwhile, Dr Chen continued, genomic test panels are becoming Food and Drug Administration (FDA) approved. The Oncomine Diagnostic Target test for lung cancer considers 23 biomarkers and the Foundation Companion Diagnostic considers 324 biomarkers, to name a few.

Switching gears to the payer perspective, Dr Chen referenced the fact that payers are relying on rapidly shifting guidelines for coverage decisions for biomarkers. Among the most important factors influencing payer coverage decisions for biomarkers are comparative effectiveness data, mandates in therapeutic FDA labeling, and clinical validity. As a result of this trend, there has been a rapid influx of drugs and new data. There are over 850 targeted compounds in development, Dr Chen referenced, and 3200 active clinical trials supporting new drug development. Over 500 drug companies are developing targeted therapies, he added.

However, in order to be able to utilize these therapies, a better understanding of the targets are needed, Dr Chen argued.

He then posed the question of the right way to select and administer treatments in precision medicine. To this end, there are five rights of precision medicine-based treatment (right diagnosis, right target right treatment, right monitoring, and right testing) and five data needs of precision medicine (diagnostic biomarker, prognostic biomarker, predictive biomarker, monitoring biomarker, and right testing).


Related Content

Precision Medicine Improves Survival in Late-Stage Cancer, Lowers Health Care Costs

Survey: Precision Medicine Program a “High Priority” in Oncology Health Care Systems


There are multiple challenges accompanying precision medicine data, Dr Chen explained, including heterogeneous biomarkers, targets and drugs; rapidly updating guidelines and biomarkers; difficulty in tracking; discerning which biomarkers are true and appropriate; and which are cost-effective. These challenges form the commonly referenced “Five V” challenges—variety, velocity, volume, veracity, and value, respectively—associated with big data. In this regards, Dr Chen told Journal of Clinical Pathways that “Precision medicine is a big data problem. [Oncologists] need to think of it in this light.”

In his other concluding points, Dr Chen stressed that molecular tumor boards should only be utilized in exceptional cancer cases, there should be a clearer distinction between precision medicine and genomics centered around biomarkers and other factors, traditional algorithms are hampered by insufficient and incomplete data and should focus on finding optimal patient matches rather than optimal biomarkers, and similarity algorithms may permit rapid data reuse.

“Data capture and sharing among oncologists are essential to move progress in precision cancer medicine,” he concluded.—Zachary Bessette