The GO Tumor Board was designed to support a fully integrated workflow from case creation to recommendations report sign-off by an oncologist and, in conjunction with other GO tools, for large scale analysis of data analyzed through the application.
- Patient history is captured in the timeline feature (screenshot, below), which allows for input of the patient’s chronological medical history including all prior diagnoses, clinical events (e.g., biopsy, relevant past medical history) and treatments. This timeline can be displayed as a graphical image or as an event table.
- As shown below, all markers that are input into the application (either via a VCF, parsed from XML or manually added) are annotated with all available information about the gene from My Cancer Genome in the marker details section. For NGS variants, the tool has a proximity lookup table which allows users to see similar variant entries in COSMIC and aids users in determining the potential pathogenicity of VUSs. Finally, for the variants themselves, there are a number of database annotations available including population frequencies (1000 genomes, ExAC, EVS), COSMIC lookup, and prediction scores (SIFT, Mutation Taster). Annotations data can be extended using any internal data or in-licensed data to which institutions like UK Markey have access rights (we can create loaders for proprietary data sources, and have done this for several of our clients). Annotations can also be added manually while using this Tumor Board app.
- Relevant therapies for the patient’s tumor type(s) and/or for other tumor type(s) are generated based on patient information. These recommendations can be selected for inclusion on the final Tumor Board report/presentation. These therapy and prognosis recommendations are populated from FDA, NCCN, ASCO and MCG guidelines contained within the GO KMS. Therapies that have a source of FDA indicate that that therapy is recommended on-label by the FDA in the package insert. Therapy recommendations can be separated within the application’s user interface to differentiate FDA-recommended vs. off-label therapies.
- Clinical trial recommendations are sourced from clinicaltrials.gov, and cancer.gov trials are curated by the team at My Cancer Genome on a weekly basis. All relevant trials that match based on the conditions of a case are presented in the recommendations section of the application and can be toggled on/off for inclusion in the final recommendation/report. At present, there are >2400 fully biomarker-driven trials in our system that have been curated by the My Cancer Genome team.
- The system may be configured to display institution-specific clinical trials, including their internal names and status information, before any clinical trials available at other locations. We are currently working with the IT team at Vanderbilt to integrate with OnCore through its APIs.
- We are in the process of extending the GO Clinical Trial model in collaboration with Vanderbilt. We have added the fields necessary for clinical trial pre-screening to the KMS data model (i.e., those used in trial inclusion and exclusion criteria) to allow for the filtering of trials based on the clinical features of the patient. As we continue to extend this capability, it will be made available for your implementation.
- Literature references can be attached to recommendations in the application interface or in the backend KMS. PubMed and NCT IDs, as well as images (including full-length PDFs), can be added to each literature record.
- Tumor board decisions are captured as discrete data and are converted into editable text reports. A “Recommendations” report generation is based on templates that are fully customizable by client users, but GenomOncology will assist with this process during the implementation. This flexibility allows for the possibility of many types of reports to be generated from the data in tumor board (e.g., a simple clinical recommendation or a detailed report on the meaning of the molecular markers in the case, including information on each marker, images, timelines, and the recommended course of action.). Some clients choose to utilize a different template for Breast Tumor Boards vs. Lung, and so on.