Background Immunotherapy with the chimeric anti-GD2 monoclonal antibody dinutuximab, combined with alternating granulocyte-macrophage colony-stimulating factor and intravenous interleukin-2 (IL-2), improves survival in patients with high-risk neuroblastoma. We aimed to assess event-free survival after treatment with ch14.18/CHO (dinutuximab beta) and subcutaneous IL-2, compared with dinutuximab beta alone in children and young people with high-risk neuroblastoma. Methods We did an international, open-label, phase 3, randomised, controlled trial in patients with high-risk neuroblastoma at 104 institutions in 12 countries. Eligible patients were aged 1-20 years and had MYCN-amplified neuroblastoma with stages 2, 3, or 4S, or stage 4 neuroblastoma of any MYCN status, according to the International Neuroblastoma Staging System. Patients were eligible if they had been enrolled at diagnosis in the HR-NBL1/SIOPEN trial, had completed the multidrug induction regimen (cisplatin, carboplatin, cyclophosphamide, vincristine, and etoposide, with or without topotecan, vincristine, and doxorubicin), had achieved a disease response that fulfilled prespecified criteria, had received high-dose therapy (busulfan and melphalan or carboplatin, etoposide, and melphalan) and had received radiotherapy to the primary tumour site. In this component of the trial, patients were randomly assigned (1:1) to receive dinutuximab beta (20 mg/m(2) per day as an 8 h infusion for 5 consecutive days) or dinutuximab beta plus subcutaneous IL-2 (6 x 10(6) IU/m(2) per day on days 1-5 and days 8-12 of each cycle) with the minimisation method to balance randomisation for national groups and type of high-dose therapy. All participants received oral isotretinoin (160 mg/m(2) per day for 2 weeks) before the first immunotherapy cycle and after each immunotherapy cycle, for six cycles. The primary endpoint was 3-year event-free survival, analysed by intention to treat. This trial was registered with ClinicalTrials.gov, number NCT01704716, and EudraCT, number 2006-001489-17, and recruitment to this randomisation is closed. Findings Between Oct 22, 2009, and Aug 12, 2013, 422 patients were eligible to participate in the immunotherapy randomisation, of whom 406 (96%) were randomly assigned to a treatment group (n=200 to dinutuximab beta and n=206 to dinutuximab beta with subcutaneous IL-2). Median follow-up was 4.7 years (IQR 3.9-5.3). Because of toxicity, 117 (62%) of 188 patients assigned to dinutuximab beta and subcutaneous IL-2 received their allocated treatment, by contrast with 160 (87%) of 183 patients who received dinutuximab beta alone (p<0.0001). 3-year event-free survival was 56% (95% CI 49-63) with dinutuximab beta (83 patients had an event) and 60% (53-66) with dinutuximab beta and subcutaneous IL-2 (80 patients had an event; p=0.76). Four patients died of toxicity (n=2 in each group); one patient in each group while receiving immunotherapy (n=1 congestive heart failure and pulmonary hypertension due to capillary leak syndrome; n=1 infection-related acute respiratory distress syndrome), and one patient in each group after five cycles of immunotherapy (n=1 fungal infection and multi-organ failure; n=1 pulmonary fibrosis). The most common grade 3-4 adverse events were hypersensitivity reactions (19 [10%] of 185 patients in the dinutuximab beta group vs 39 [20%] of 191 patients in the dinutuximab plus subcutaneous IL-2 group), capillary leak (five [4%] of 119 vs 19 [15%] of 125), fever (25 [14%] of 185 vs 76 [40%] of 190), infection (47 [25%] of 185 vs 64 [33%] of 191), immunotherapy-related pain (19 [16%] of 122 vs 32 [26%] of 124), and impaired general condition (30 [16%] of 185 vs 78 [41%] of 192). Interpretation There is no evidence that addition of subcutaneous IL-2 to immunotherapy with dinutuximab beta, given as an 8 h infusion, improved outcomes in patients with high-risk neuroblastoma who had responded to standard induction and consolidation treatment. Subcutaneous IL-2 with dinutuximab beta was associated with greater toxicity than dinutuximab beta alone. Dinutuximab beta and isotretinoin without subcutaneous IL-2 should thus be considered the standard of care until results of ongoing randomised trials using a modified schedule of dinutuximab beta and subcutaneous IL-2 are available. Copyright (c) 2018 Elsevier Ltd. All rights reserved.