Dinutuximab beta-based first-line maintenance therapy for newly diagnosed high-risk neuroblastoma following a COG-based pathway: a single-center real-world study in China

This study delineates the feasibility, cycle-stratified safety profile, and relapse pattern of dinutuximab beta-based first-line maintenance for children with newly diagnosed high-risk neuroblastoma managed under a Children’s Oncology Group (COG)-based pathway at a single Chinese center. This retrospective study included 13 children with newly diagnosed high-risk neuroblastoma who received dinutuximab beta maintenance after COG-based multimodal treatment from 2021 to 2024. A contemporaneous 43-patient cohort, including 30 who did not receive dinutuximab beta (7 received naxitamab), provided treatment-pathway and accessibility context only and was not a comparator. Maintenance-phase event-free survival (EFS) and overall survival (OS) from first infusion, end-of-maintenance disease status, and CTCAE v5.0 adverse events were summarized descriptively. Cycles were classified as immunotherapy-only or combined chemotherapy-immunotherapy; no formal testing or within-patient adjustment was performed. Before maintenance, 10/13 patients were in complete response (CR) and 3/13 in partial response (PR) with marrow-confined disease; 10 underwent autologous stem cell transplantation (ASCT). At the end of immunotherapy, 11 were classified as CR and 2 as progressive disease (PD). Median follow-up was 28.6 months, and 3-year maintenance-phase EFS was 69.2% (95% CI, 44.1–94.3). Only one death occurred, rendering OS estimates immature. Among 65 cycles (41 immunotherapy-only; 24 combined), pain, cytokine release syndrome, and capillary leak syndrome were predominantly low grade. Grade ≥ 3 myelosuppression, neutropenia, infection, and fever were numerically more frequent in combined than immunotherapy-only cycles (83.3% vs. 34.1%, 58.3% vs. 9.8%, 37.5% vs. 19.5%, and 37.5% vs. 26.8%, respectively; descriptive only). Four EFS events occurred: one pathology-confirmed parenchymal CNS relapse, one MRI-defined probable central nervous system (CNS) relapse without histologic confirmation, one non-CNS skull-base osseous progression with parameningeal extension, and one abdominal nodal progression followed by death. No treatment-related deaths occurred. Because this cohort entered maintenance with low residual disease burden after multimodal treatment, outcomes cannot be attributed to dinutuximab beta alone. Cycle-stratified reporting contextualized adverse-event patterns but did not establish component-specific causality. The three heterogeneous cranial-region events were hypothesis-generating and do not indicate an increased CNS-relapse rate or an altered relapse spectrum.