Tumor mutational burden (TMB), a measure of the number of somatic mutations per Mb of the assay target region, is becoming a common biomarker in prediction of immunotherapy response. As the tumor evolves, mutations are accumulated, leading to growth advantages and opportunities for presentation of new neoantigens. Immunotherapies designed to antagonize common checkpoint inhibitors like PD1/PD-L1 and CTLA4 correct for this immune evasion and mutational burden provides a continuous variable that’s prognostic
to this attainment. This is also showing promise in second and third line inhibitory compounds (TIM-3, LAG3, IDO1) and co-stimulatory antibodies (OX40, GITR, CD40).
Calculation of TMB is dependent on calling tumor-derived (somatic) mutations that are derived from identifying variants in the tumor cells and contrasting them against variants in a matched normal/germline sample. However, this approach increases costs of testing and in many cases, may not be possible if a normal sample cannot be obtained, consented or absent from a clinical trial protocol for TMB testing. To increase applicability of TMB to the volume of retrospective studies, calculation must be able to be performed by using only the tumor specimen and we feel our approach provides robust calculation across use cases.
Download our poster from SITC 2018 to learn more about our tumor-only TMB pipeline, using somatic classifications determined using a random forest model.