TMB Harmonization Working Group Meeting | Friends of Cancer Research

TMB Harmonization Working Group Meeting

TMB Harmonization Working Group Meeting

On May 10, 2018, Friends of Cancer Research (Friends) convened stakeholders across all health sectors to assess current methods of measuring tumor mutational burden (TMB) and develop a consensus approach to the standardization of this biomarker. This roundtable built on a workshop hosted by Friends in the Fall of 2017, where a group of experts discussed potential industry standards/guidelines for the harmonization of TMB measurement and reporting.


Project Background
The 2017 workshop focused on three components that would benefit from standardization: defining TMB, establishing an analytical validation approach, and ensuring optimal clinical implementation. It was agreed that to achieve these goals a multi-step approach was necessary, where diagnostic companies would seek agreement on analytical and clinical validation studies to support a standardized method of TMB measurement and reporting.

The three steps consist of an in silico TMB analysis using publicly available data from The Cancer Genome Atlas (TCGA), an empirical TMB analysis using publicly available human tumor cell lines, and a clinical analysis using clinical samples.


Meeting Summary
To achieve step one of the project, the May 10th Friends roundtable brought key stakeholders together to review the findings of the in silico analysis. Several diagnostic and clinical partners shared insights gained from the in silico analysis of TCGA data, highlighted considerations of the study design and statistical analysis plan for the empirical TMB study, and discussed implications and needs for the implementation and clinical validation of TMB.


Strong evidence supporting TMB as a robust biomarker was presented during the roundtable, and participants highlighted initial lessons learned, as well as opportunities for continued improvement, including:

  • Access to TMB assays and breadth of clinical application
  • Optimization of an objective cut-point (or cut-point range) for patient selection
  • Awareness of need for sufficient tissue to conduct analyses


Additionally, the need for consistency in TMB measurement and reporting was outlined and several opportunities for TMB technical harmonization were provided, such as:

  • Agreement on analytical parameters for TMB calculation
  • Generation of a universal reference standard as an alignment tool
  • Agreement on statistical approaches that will lead to consistent TMB calculation and clinical interpretation

There was consensus that this harmonization effort will lead to improving familiarity with TMB measurement and confidence on its predictive ability.


TMB Harmonization Project - Step One Summary
Eight diagnostic partners (Foundation Medicine, Inc., Guardant Health, Inc., Illumina, Inc., Memorial Sloan Kettering Cancer Center (MSKCC), NeoGenomics Laboratories, Inc., Personal Genome Diagnostics (PGDx), QIAGEN, Inc., Thermo Fisher Scientific) took part in an in silico analysis that compared the correlation between TMB scores calculated using whole exome sequencing (WES) and their own gene panels. The diagnostic partners used the same data for this analysis (the Multi-Center Mutation Calling Multi-tumor Completion effort, or MC3 TCGA data), which consisted of thousands of samples from various cancer types. The goal of this initial step of the project was, for the first time, to compare how the bioinformatic approaches to determine TMB used by eight leading tests relates to one another.


The results showed that TMB determined by the different tests is relatively consistent across a range of mutational burden. Considering the various methods employed to analyze and count mutations, the initial in silico analysis results were viewed by the participants as a positive outcome.  While variation between the different platforms does exist, the initial performance analysis shows that TMB as evaluated by the eight data partners is in a close enough range that further alignment is achievable.


Sources of variation between panels, such as filter thresholds, variant types included, and genome coverage, were discussed. Based on this initial analysis, several suggestions were proposed that would improve the analytical validation of this novel biomarker. The final in silico TMB data will be used for the development of a manuscript to be submitted for publication at a peer-reviewed journal.


Next Steps
The empirical analysis—step two in the analytical validation effort—will use findings from the in silico analysis to inform the bioinformatics approach for the calculation of TMB using WES of human tumor-derived cell lines. The cell lines selected will represent a broad array of clinically-meaningful TMB values and have matched normal tissue. The working group will agree on the selection of the cell lines and a central laboratory to handle them and create the reference standard. Several statistical approaches for alignment of panel-derived TMB values to the reference standard were also discussed at the meeting. The reference standard will be used as an alignment tool to make TMB values calculated using different gene panels comparable to one another, thus providing consistency during TMB reporting. 


At the end of the roundtable, meeting participants discussed the clinical utility of TMB and outlined several clinically-relevant considerations for the implementation of TMB in clinical decision-making, including physician education, tissue availability and amount, and the potential use of blood-based TMB assays.


Friends will continue to lead this harmonization effort to ensure patients have access to reliable and consistent testing for biomarkers that guide critical treatment decisions.


For more information on this project please Click HERE

Return to blog
Back to Top