Design and Implementation of a Decentralized Virtual Molecular Tumor Board across a Large, Diverse Community Health System

Document Type


Publication Date



oregon; chiles; washington; swedish


Introduction: Rapid advances in precision cancer therapy and immunotherapy are transforming oncology with a wealth of genomic biomarker-associated drugs recently approved or available through clinical trials. Concomitantly, clinical tumor genomic sequencing has expanded in scope and utilization. As such, oncologists are now faced with treatment decisions of increasing complexity that may span conventional, precision therapies and immunotherapies for a single patient. However, typically supporting tumor board infrastructures are often suboptimal, with isolated boards carried out at individual sites and not utilizing large-scale analytics. Here we describe our efforts in designing and implementing a decentralized virtual molecular tumor board (VMTB) across a large health system spanning 50+ hospitals across seven states. Methods: A software platform was developed that integrates clinical genomics datasets, digital pathology imaging, and radiology imaging, as well as restructures key elements from the electronic health records. The platform was deployed in a secure internal cloud environment at Providence and made available systemwide to all clinicians. VMTBs were carried out twice a month via videoconference. Over the initial six-month launch window of the VTMB study, data were manually abstractsed for aggregate analysis. Results: Key novel features integrated in the platform include real-time clinical trial matching, deep integration of medical imaging, timeline visualizations of a patient’s treatment trajectory, and treatment comparisons via patient matching across the health system. Over the pilot phase of the VMTB, the most common tumor types discussed at the conference were of brain origin, followed by colon, lung, and bile duct. For patients discussed at the VMTB, 62% had a recommendation for treatment with a precision therapy or immunotherapy, either standard of care or through a clinical trial enrollment. This is in comparison to historical utilization of 18% for precision therapies, 30% for immunotherapies, and 4% for clinical trial enrollment for genomicstested patients that were not presented at the VMTB. Conclusions: The VMTB developed into a forum for multidisciplinary experts across a broad geographic region to assist with critical treatment decisions for cancer patients. Patients discussed at the VMTB were more likely to receive a precision therapy or immunotherapy. Although part of this is likely due to selection bias (i.e., cases without actionable findings would be less likely to be presented at a molecular tumor board), we also expect the knowledge gained by participating physicians to lead to appropriate increases in precision therapy or clinical trial utilization over the larger patient population.

Clinical Institute





Earle A. Chiles Research Institute


Presented at the AMP Annual Meeting; November 1-5; Phoenix, AZ. 2022: O001.

This document is currently not available here.