Risk stratification by somatic mutation burden in Ewing sarcoma.

Abstract

Up to one-third of patients with localized Ewing sarcoma (ES) develop recurrent disease, but current biomarkers do not accurately identify this high-risk group. Therefore, the objective of this study was to determine the utility of mutational burden in predicting outcomes in patients with localized ES. Clinical and genomic data from 99 patients with ES, of whom 63 had localized disease at diagnosis, were obtained from the cBioPortal for Cancer Genomics. Genomic data included the type and number of somatic mutations using cBioPortal mutation calling. Primary endpoints were overall survival (OS) and the time to progression (TTP). Patients had a median number of 11 somatic mutations. Patients were stratified according to whether they had a lower or higher mutational burden if they had ≤11 or >11 mutations, respectively. Higher mutational burden was significantly associated with inferior OS and TTP, a finding that was confirmed by univariate and multivariable analyses. In patients who had localized disease at diagnosis, higher mutational burden was the only variable significantly associated with inferior OS and TTP. The presence of a mutation in either stromal antigen 2 (STAG2) or tumor protein 53 (TP53), both of which were correlated previously with shorter OS in patients with ES, were significantly associated with higher mutational burden. Upon stratifying patients who had localized disease based on a standard panel of cancer genes, higher risk stratification was correlated significantly with inferior TTP and trended toward significance with inferior OS. Patients who have localized ES and a higher mutational burden have inferior OS and TTP compared with those who have lower mutation burden. The current findings suggest that the somatic mutation burden can be used to better risk stratify these patients and to guide clinical decision making.