How to Run an Effective Tumor Board

How to Run an Effective Tumor Board

A guide to the roles, processes, and data management

Marissa Schuh, Rachel Stewart, DO, PhD, Riham El Khouli, MD, PhD, Rachel W. Miller, MD, Justine Pickarski, MS, LGC, Eric B. Durbin, DrPH, MS, Susanne Arnold, MD, Jill Kolesar, PharmD, MS

Oncology is unique in that somatic mutations can both drive the development of a tumor and serve as a therapeutic target for treating the cancer. Clinical sequencing of many cancers is now a routine part of care; however, the volume and complexity of sequencing data generated can be overwhelming to practicing oncologists. Molecular tumor boards (MTBs), typically multidisciplinary and disease agnostic, have emerged as a method for analyzing sequencing data and providing clinically relevant recommendations to the treating physician. A number of factors contribute to the effective implementation of an MTB.

Key considerations

MTBs should evaluate their scope by considering the needs of the constituency, the population, and the institution, and the research interests of the team. The focus should be on providing the highest-quality analysis, with broad intellectual support across multiple disciplines. Creating a set of guidance documents and levels of evidence for decision making will place all decisions into context, allowing individual practitioners who receive the MTB reviews to easily understand the data that support recommendations. These measures also give third-party payers and institutions confidence in the decisions of the MTB, and mediate medicolegal risks, while supporting the provision of targeted therapy by insurers.

Members and roles

Identifying the leadership team of an MTB is a critical first step. At our institution, the University of Kentucky Markey Cancer Center, the co-leaders are a clinical pharmacologist and a gynecologic oncologist. The co-leaders have complementary expertise in the clinical care of cancer patients and the molecular pharmacology of anticancer agents, and provide scientific, administrative, and clinical leadership to the MTB.

A pathologist with board certification and/or subspecialty expertise in molecular genetic pathology should be included on an MTB in order to assist with variant interpretation, discussion of clinically actionable variants, and questions regarding test methodology and quality assurance practices. Input from a pathologist can also be useful for prioritizing molecular testing in the setting of small or limited tumor samples. The clinical significance of variants and resulting treatment recommendations are best discussed within the context of the primary site and histopathologic diagnosis. A pathologist with expertise in surgical pathology can add valuable information to the discussion by presenting histopathologic findings, staging information, and interpreting the results of immunohistochemical stains and other ancillary tests.

Patients referred for MTB consultation are typically refractory or relapsing patients who have exhausted all clinical standard of care options, which makes review of imaging a critical component of the process. Board-certified, fellowship-trained radiologists are key members of the MTB and provide expertise regarding disease extent and progression. Careful review and comparison of all prior imaging studies using the standard treatment response metrics—including the Response Evaluation Criteria in Solid Tumors (RECIST 1.1) and Positron Emission Response Criteria in Solid Tumors (PERCIST 1.0)—is valuable. Additionally, selection of representative images for presentation during the MTB session helps the group understand the disease behavior and the multidisciplinary assessment of each patient.

Our clinical team is composed of physicians, clinical pharmacologists, and genetic counselors. Input from surgical and medical oncologists with broad clinical expertise for the most common tumor types seen by the MTB, and with experience in therapeutic decision making in the era of targeted therapies, is essential. Clinical trialists are also valuable, as they not only review cases for clinical recommendations but also identify potential candidates for clinical trials with biomarker entry criteria. Clinical pharmacologists provide a critical assessment of the drugability of identified mutations, potential treatment strategies, and pharmacological differences within a drug class. A genetics counselor is also present at all MTB meetings to identify candidates for germline testing. Our clinical team reviews the clinical case history, evaluates the targetability of identified mutations, and develops a patient-specific recommendation for either standard of care therapy, off-label therapy, a clinical trial, or additional testing.

The MTB process

Overview of the University of Kentucky Markey Cancer Center’s MTB process, including the steps that occur before, during, and after the MTB meeting. Click to see full-size.

The MTB manager oversees all aspects of MTB operations [see flowchart on next page]. The manager uses standard operating procedures to set the meeting agenda, disseminate cases to reviewers, return recommendations to treating physicians, develop documentation, and provide continuing medical education. Guidance documents serve to objectify the decisions of the group, maintain consistency, and provide structure to the review process.

Prior to the initiation of the MTB, we received approval of our initial guidance documents by the university’s policy and administrative bodies. We continuously review the guidance documents as the science changes to incorporate new findings, new technologies, and new therapeutic options for patients. Levels of evidence do not change, however, and that is the context in which we report all results.

Data management

MTB data must be properly managed to support clinical decision making, to enable the evaluation of the MTB’s performance, and to support downstream research. At a minimum, an MTB requires the ability to track patients through the MTB process. A data management system should capture patient demographics, consents, details about the cancer diagnosis and stage, prior treatments and responses, and clinically relevant variants, as well as the final recommendations made by the MTB. Information about disease progression and response to treatment post-MTB review are also critical for evaluating the impact of the MTB on patient outcomes. Our center utilizes a clinical trials management system for this purpose.

The way molecular data is reported can impact its usefulness. While the PDF reports generated by molecular testing laboratories are sufficient for clinician review, they are not suited for further analyses such as calculating variant allele frequencies and identifying retrospective cohorts or those needed for research. It is essential to obtain reported variant data (including both significant variants and variants of unknown significance) in a structured format that can be loaded into a query-able database. While no national standards yet exist for structured molecular reporting, many vendors offer reports in an eXtensible Markup Language (XML) or JavaScript Object Notation (JSON) format for this purpose. In addition to the report data, it is important to also obtain the raw sequencing data and metadata that describes the bioinformatics pipelines utilized for the samples. Our center routinely obtains raw data files in the Binary Alignment Map (BAM) format. We store these rather large files on a secure university supercomputer and in public cloud-based resources.

Molecular data is most useful for analyses and research when integrated with other clinical, biomarker, and multiomics data sources. For example, cancer registry data is often a rich source of standardized clinical, treatment, and long-term outcome information for patients. Our center has partnered with the Kentucky Cancer Registry, a National Cancer Institute Surveillance, Epidemiology and End Results (SEER) registry, to build a Cancer Research Data Commons that integrates data from the MTB, SEER registry, electronic pathology reports, clinical trials, biorepositories, and other multi-omics sources as a resource available to clinicians and researchers. A portal has been developed that allows authorized investigators to review patient populations using a combination of data points from the various sources. This provides an excellent source of preliminary data to further enhance research efforts at our academic medical center.

MTBs are an effective way to interpret and synthesize complex clinical sequencing reports, provide actionable treatment recommendations to treating physicians, and develop a database for ongoing research questions. Implementing an effective MTB requires the input and commitment of a large interdisciplinary team, strong administrative support, adoption of standardized workflows, and incorporation of evidence into clinical decision making.

Marissa Schuh

Marissa Schuh is the project manager of the Markey Cancer Center Precision Medicine Center;

Rachel Stewart, DO, PhD

Rachel Stewart, DO, PhD, is an assistant professor of pathology and the scientific associate director of the Markey Biospecimen Core

Riham El Khouli, MD, PhD

Riham El Khouli, MD, PhD, is an assistant professor of radiology and the medical director of research in the department of radiology at the University of Kentucky

Rachel W. Miller, MD

Rachel W. Miller, MD, is an associate professor of obstetrics and gynecology and co-director of the Markey Cancer Center Molecular Tumor Board

Justine Pickarski, MS, LGC

Justine Pickarski, MS, LGC, is a genetic counselor supervisor at the Markey Cancer Center

Eric B. Durbin, DrPH, MS

Eric B. Durbin, DrPH, MS, is an assistant professor of biomedical informatics and director of the Kentucky Cancer Registry and the Cancer Research Informatics Shared Resource Facility

Susanne Arnold, MD

Susanne Arnold, MD, is a professor of medicine and associate director of clinical translation at the Markey Cancer Center;

Jill Kolesar, PharmD, MS

Jill Kolesar, PharmD, MS, is a professor of pharmacy and co-director of the Markey Cancer Center Molecular Tumor Board.