Purpose: Although multimodal chemotherapy has improved outcomes for patients with osteosarcoma, the prognosis for patients who present with metastatic and/or recurrent disease remains poor. In this study, we sought to define how often clinical genomic sequencing of osteosarcoma samples could identify potentially actionable alterations.Experimental Design: We analyzed genomic data from 71 osteosarcoma samples from 66 pediatric and adult patients sequenced using MSK-IMPACT, a hybridization capture-based large panel next-generation sequencing assay. Potentially actionable genetic events were categorized according to the OncoKB precision oncology knowledge base, of which levels 1 to 3 were considered clinically actionable.
Results: We found at least one potentially actionable alteration in 14 of 66 patients (21%), including amplification of CDK4 (n = 9, 14%: level 2B) and/or MDM2 (n = 9, 14%: level 3B), and somatic truncating mutations/deletions in BRCA2 (n = 3, 5%: level 2B) and PTCH1 (n = 1, level 3B). In addition, we observed mutually exclusive patterns of alterations suggesting distinct biological subsets defined by gains at 4q12 and 6p12-21. Specifically, potentially targetable gene amplifications at 4q12 involving KIT, KDR, and PDGFRA were identified in 13 of 66 patients (20%), which showed strong PDGFRA expression by IHC. In another largely nonoverlapping subset of 14 patients (24%) with gains at 6p12-21, VEGFA amplification was identified.
Conclusions: We found potentially clinically actionable alterations in approximately 21% of patients with osteosarcoma. In addition, at least 40% of patients have tumors harboring PDGFRA or VEGFA amplification, representing candidate subsets for clinical evaluation of additional therapeutic options. We propose a new genomically based algorithm for directing patients with osteosarcoma to clinical trial options.
©2019 American Association for Cancer Research.