10018 Background: All patients with Ewing sarcoma (EWS) currently receive the same intensive chemotherapy and novel risk-adapted approaches to therapy are needed. We have previously shown that elevated baseline ctDNA burden is prognostic among patients with EWS. Here we investigate the prognostic impact of on-treatment ctDNA burden in patients with metastatic EWS treated on COG AEWS1221. Methods: AEWS1221 was a randomized phase 3 trial evaluating ganitumab plus interval compressed chemotherapy (vincristine/doxorubicin/cyclophosphamide alternating with ifosfamide/etoposide given every 2 weeks) for patients with newly diagnosed metastatic EWS. We evaluated 680 serial blood samples and data from 273 patients enrolled to AEWS1221. We utilized a custom NGS hybrid-capture assay for ctDNA quantification that detects EWSR1 and FUSfusions. We used log rank tests to compare event-free survival (EFS) from the designated timepoint according to ctDNA burden post cycle 1 of VDC (planned day 14) and post cycle 2 VDC+IE (planned day 28) of chemotherapy. We dichotomized patients by ≥0.5% ctDNA at each timepoint. A P-value of ≤0.05 was considered evidence of a significant association. Results: Of 298 eligible patients with metastatic EWS, 273 provided blood samples for ctDNA analysis. Three patients had non-FET-ETS family fusions ( CIC::DUX4, EWSR1::WT1 and EWSR1::GTDC1) and were excluded from further analysis. Eighty-five percent of patients were < 21, 56% were male, 33% had pelvic bone primary tumors, and 40% had lung only metastatic disease. ctDNA was ≥0.5% in 86% of patients at baseline (median 25%, range 0-113%, n = 264), in 53% (median 0.53%, range 0-15%, n = 34) after 1 cycle (median day 14, range 11-20), and in 14% (median 0, range 0-53%, n = 130) after 2 cycles (median day 29, range 25-36). Baseline ctDNA burden ≥0.5% was associated with an increased risk of EFS-event (3-year EFS 35% [95% CI 29-41%] vs. 61% [95% CI 43-76%], HR = 2.2, P= 0.002). ctDNA ≥0.5% after cycle 1 was also associated with an increased risk of EFS-event (3-year EFS 6% [95% CI 0.4-22%] vs. 56% [95% CI 29-76%], HR = 6.0, P< 0.0001). ctDNA ≥0.5% after cycle 2 was not associated with an increased risk of EFS-event (28% [95% CI 10-49%] vs. 34% [95% CI 25-43%], HR = 1.2, P= 0.55). Conclusions: ctDNA burden ≥0.5% following one cycle of chemotherapy identifies patients highly likely to relapse, albeit in a small cohort with available data at that timepoint. Conversely, ctDNA at a cutpoint of ≥0.5% after two cycles of chemotherapy was not prognostic, suggesting either more sensitive assays are needed or the prognostic value of ctDNA burden is diminished following additional therapy. These findings will enable novel trials of risk-adapted therapy focused on baseline and early ctDNA burden.