Survival outcomes in pediatric patients with metastatic Ewing sarcoma who achieve a rapid complete response of pulmonary metastases

11535 Background: Twenty-five percent of patients with Ewing Sarcoma (EWS) have metastatic disease at diagnosis, most commonly in the lungs. Our objectives were to compare overall (OS) and pulmonary disease-free survival (PDFS) between patients with metastatic EWS at diagnosis who achieve rapid complete response (RCR), defined as radiographic resolution of all pulmonary nodules at the end of induction chemotherapy, and those with residual pulmonary nodules after induction chemotherapy (non-RCR). Methods: This multi-center, retrospective cohort study included children under 20 years of age with metastatic EWS treated from 2007-2020 at 19 institutions participating in the Pediatric Surgical Oncology Research Collaborative. Chi-square tests were conducted for differences among groups, and p < 0.05 was considered significant. Kaplan-Meier curves were generated for OS and PDFS. Cox regression was performed for OS, controlling for whole lung irradiation (WLI), extrapulmonary metastases, and age. Results: Among 153 patients with metastatic EWS at diagnosis, 61 (40%) achieved RCR, and 87 (57%) did not (5 unknown). The median number of pulmonary nodules was 3.5 in the RCR vs 5 in the non-RCR groups (p = 0.02), while bilateral disease was present in 44 (72%) RCR vs 75 (86%) non-RCR patients (p = 0.03). Pulmonary relapse occurred in 58 (38%) patients, including 18 (29%) in the RCR and 36 (41%) in the non-RCR groups (p = 0.14). Factors associated with pulmonary relapse in the univariate analysis include median age [15 years (IQR: 12-17) with pulmonary relapse vs 12 years (IQR: 9-16) without (p = 0.03)] and extrapulmonary metastases at diagnosis [24 (41%) patients with pulmonary relapse vs 26 (27%) patients without pulmonary relapse (p = 0.05)]. WLI was performed in 100 patients and was not associated with a reduction in pulmonary relapse (p = 0.93). Complete surgical clearance of pulmonary disease was attempted in 29 (19%) patients, including 17 of the patients who met the criteria for RCR. On Kaplan-Meier analyses, 5-year PDFS did not significantly differ based on RCR (67%) vs. non-RCR (53%, p = 0.13), or WLI (61%) vs. no WLI (54%, p = 0.32). However, 5-year OS did vary significantly between patients with RCR and WLI (85%), RCR without WLI (55%), non-RCR with WLI (60%), and non-RCR without WLI (25%; p < 0.01). In Cox regression analysis, RCR was associated with improved OS after controlling for WLI, extrapulmonary metastases, and patient age (HR 0.51, p = 0.03). Conclusions: Patients with EWS who had pulmonary metastases at diagnosis had improved OS if they achieved RCR and received WLI, despite having no significant differences in rates of pulmonary relapse. Further research is needed to determine whether clearance of pulmonary disease achieved through surgical metastasectomy provides a comparable survival benefit and if any subsets of patients have a reduced risk of pulmonary relapse after resection.