Does Longitudinal Targeted Panel Sequencing Provide Clinically Relevant Information in Translocation-Associated Sarcomas?

Abstract

Translocation-associated sarcomas (TAS) represent a heterogeneous group of pathologic entities characterized by specific gene fusions and a low tumor mutational burden (TMB). Although RNA sequencing remains the hallmark molecular platform to establish the fusion type, targeted panel DNA next-generation sequencing (NGS) has been increasingly used in TAS to identify actionable alterations as well as potential mechanisms of resistance to therapy. We herein investigate sequential targeted DNA sequencing (Memorial Sloan Kettering Cancer Center-Integrated Mutation Profiling of Actionable Cancer Targets) in a large cohort of 108 TAS, spanning six most common subtypes, with available longitudinal sequencing. Overall, 53% of cases acquired alterations (n = 122) compared with the initial testing, of which 33 (27%) were oncogenic/likely oncogenic. Only two of the five patients with >3 tests/each acquired oncogenic mutations. Overall, 21% of cases did not acquire alterations in any assay. Subsequent NGS was initiated after systemic treatment (93%) and disease progression (83%). The number of oncogenic acquired alterations did not correlate with histotype (mean 0.4/case). In sequential NGS, the mean mutation count, TMB, and fraction of genome altered were significantly higher (<i>P</i> < .001, .04, and .03, respectively) compared with baseline. <i>TP53</i> alteration was the most common acquired mutation (10%), being detected mostly in the subsequent assay. By contrast, other common alterations, such as <i>STAG2</i> in Ewing sarcoma and gene amplifications in alveolar rhabdomyosarcoma, were detected at diagnosis. Although 27% of pathogenic alterations in the initial and 9% in the subsequent NGS were deemed as OncoKB levels 1-2 for other cancers, none were relevant for sarcomas. The role of multiple NGS testing in the context of TAS remains undetermined.