(P 49) A WORKING GROUP FOR THE DEVELOPMENT OF A CLINICAL FRAMEWORK THAT STRATIFIES OSTEOSARCOMA PATIENTS USING BIOMARKERS OF CLINICAL RISK AND BIOLOGY

Objective: Despite the well-appreciated extensive heterogeneity of osteosarcoma, all patients diagnosed with this disease receive the same initial treatment. Clinical risk and prognostication have, until now, focused on the presence of metastatic disease at diagnosis and histologic response to chemotherapy. Our approach to treatment has not yet incorporated molecular features, although several recent advances propose biomarkers that could identify subgroups of patients with distinct clinical risk or potential to respond to specific therapies. Therefore, there is a critical need to develop a harmonized approach to the biological and risk stratification of patients with osteosarcoma based on molecular features, incorporating new technologies and recent discoveries.

Methods: To evaluate potential molecular prognostic biomarkers, we organized an international working group of disease experts to assess the most promising and mature biomarker candidates. We reviewed the prognostic significance of contemporary molecular biomarkers with strong pre-clinical evidence, including MYC amplification, other copy number variants, circulating tumor DNA (ctDNA), tumor microenvironment composition, and multi-omic classification strategies. We evaluated the current level of evidence for each biomarker and its potential to inform treatment, with additional consideration for the practicalities of incorporation into the current, well-established regimens. Through this process, we identified molecular features that this panel recommends for assessment in all future phase II/III studies.

Results: Preliminary evidence supports copy number changes (particularly MYC amplification) as promising prognostic biomarkers in osteosarcoma. Several studies demonstrate ctDNA burden is associated with outcomes independent of known risk factors. Additional potential biomarkers with prognostic significance include tumor microenvironment composition, genome-wide loss of heterozygosity, multiple candidate transcriptomic signatures, proteomic and epigenetic frameworks for osteosarcoma classification, and combinations of these approaches. These markers require further validation, determination of biological overlap, integration with clinical features associated with outcome, and the development of standardized, clinically tractable assays that facilitate clinical implementation.

Conclusion: These biomarkers, in combination with clinical risk factors, have the potential to identify clinically and biologically relevant osteosarcoma subgroups. Incorporating this framework into clinical care will require prospective validation in upcoming trials, but has significant potential to inform our development of risk- and biology-based therapeutic strategies that leverage our growing understanding of the biology of this disease. Our review highlights the need for comprehensive evaluation of these biomarkers in large, annotated cohorts with pre-planned analysis so that the prognostic impact of each marker can be definitively determined in the context of relevant clinical features.  These studies will provide a platform for realizing risk- and biology-adapted treatment strategies for osteosarcoma.