ctDNA quantification improves estimation of outcomes in patients with high-grade osteosarcoma: a translational study from the OS2006 trial

Purpose Osteosarcoma stratification relies on clinical parameters and histological response. We developed a new personalized stratification using less invasive circulating tumor DNA (ctDNA) quantification. Patients and Methods Plasma from patients homogeneously treated in the prospective protocol OS2006, at diagnosis, before surgery and end of treatment, where sequenced using low passage whole genome sequencing (lpWGS) for copy number alteration detection. We developed a prediction tool including ctDNA quantification and known clinical parameters to estimate patients’ individual risk of event. Results CtDNA quantification at diagnosis (diagCPA) was evaluated for 183 patients of the protocol OS2006. diagCPA as a continuous variable was a major prognostic factor, independent from other clinical parameters, including metastatic status (diagCPA HR=3.5, p=0.002 and 3.51, p=0.012, for PFS and OS). At the time of surgery and until the end of treatment, diagCPA was also a major prognostic factor independent from histological response (diagCPA HR=9.2, p<0.001 and 11.6, p<0.001, for PFS and OS). Therefore, the addition of diagCPA to metastatic status at diagnosis or poor histological response after surgery improved the prognostic stratification of patient with Osteosarcoma. We developed the prediction tool PRONOS to generate individual risk estimations, showing great performance with . ctDNA quantification at the time of surgery and the end of treatment still required improvement to overcome the low sensitivity of lpWGS and to enable the follow up of disease progression. Conclusions The addition of ctDNA quantification to known risk factors improves the estimation of prognosis calculated by our prediction tool PRONOS. To confirm its value an external validation in the Sarcoma 13 trial is underway.