To explore the effects of immunotherapy in the International Society of Paediatric Oncology Europe Neuroblastoma Group SIOPEN high-risk neuroblastoma 1 trial (HR-NBL1 trial), two cohorts were studied: one prior to and one after the introduction of dinutuximab beta. All patients received standard induction and high-dose therapy (HDT) with autologous stem cell rescue (ASCR); the local control comprised surgery and radiotherapy to the primary tumour site, followed by isotretinoin. A landmark timepoint of 109 days, resulting from the median time between ASCR and initiation of immunotherapy, was used to define patients’ eligibility in the pre-immunotherapy analysis cohort. Median follow-up was 5.8 years (inter-quartile range (IQR): 4.2–8.2 years) for 844 eligible patients balanced for risk factors, such as age, sex, stage 4, MYCN amplification and response prior to HDT. The five-year event-free and overall survival (95% confidence interval (CI) of 466 patients not receiving immunotherapy was 42% (38–47%) and 50% (46–55%) but was 57% (51–62%) and 64% (59–69%) for 378 patients receiving immunotherapy (p < 0.001). A multivariate analysis identified absence of immunotherapy (p = 0.0002, hazard ratio (HR) 1.573); type of HDT (p = 0.0029, HR 1.431); less than complete response prior to maintenance therapy (p = 0.0043, HR 1.494) and >1 metastatic compartment at diagnosis (p < 0.001, HR 2.665) as risk factors for relapse or progression. Results suggest an important role for dinutuximab beta-based immunotherapy within the treatment concepts applied in HR-NBL1/SIOPEN.

Tumor segmentation is one of the key steps in imaging processing. The goals of this study were to assess the inter-observer variability in manual segmentation of neuroblastic tumors and to analyze whether the state-of-the-art deep learning architecture nnU-Net can provide a robust solution to detect and segment tumors on MR images. A retrospective multicenter study of 132 patients with neuroblastic tumors was performed. Dice Similarity Coefficient (DSC) and Area Under the Receiver Operating Characteristic Curve (AUC ROC) were used to compare segmentation sets. Two more metrics were elaborated to understand the direction of the errors: the modified version of False Positive (FPRm) and False Negative (FNR) rates. Two radiologists manually segmented 46 tumors and a comparative study was performed. nnU-Net was trained-tuned with 106 cases divided into five balanced folds to perform cross-validation. The five resulting models were used as an ensemble solution to measure training (n = 106) and validation (n = 26) performance, independently. The time needed by the model to automatically segment 20 cases was compared to the time required for manual segmentation. The median DSC for manual segmentation sets was 0.969 (±0.032 IQR). The median DSC for the automatic tool was 0.965 (±0.018 IQR). The automatic segmentation model achieved a better performance regarding the FPRm. MR images segmentation variability is similar between radiologists and nnU-Net. Time leverage when using the automatic model with posterior visual validation and manual adjustment corresponds to 92.8%.

Objectives. To externally validate and assess the accuracy of a previously trained fully automatic nnU-Net CNN algorithm to identify and segment primary neuroblastoma tumors in MR images in a large children cohort. Methods. An international multicenter, multivendor imaging repository of patients with neuroblastic tumors was used to validate the performance of a trained Machine Learning (ML) tool to identify and delineate primary neuroblastoma tumors. The dataset was heterogeneous and completely independent from the one used to train and tune the model, consisting of 300 children with neuroblastic tumors having 535 MR T2-weighted sequences (486 sequences at diagnosis and 49 after finalization of the first phase of chemotherapy). The automatic segmentation algorithm was based on a nnU-Net architecture developed within the PRIMAGE project. For comparison, the segmentation masks were manually edited by an expert radiologist, and the time for the manual editing was recorded. Different overlaps and spatial metrics were calculated to compare both masks. Results. The median Dice Similarity Coefficient (DSC) was high 0.997; 0.944–1.000 (median; Q1–Q3). In 18 MR sequences (6%), the net was not able neither to identify nor segment the tumor. No differences were found regarding the MR magnetic field, type of T2 sequence, or tumor location. No significant differences in the performance of the net were found in patients with an MR performed after chemotherapy. The time for visual inspection of the generated masks was 7.9 ± 7.5 (mean ± Standard Deviation (SD)) seconds. Those cases where manual editing was needed (136 masks) required 124 ± 120 s. Conclusions. The automatic CNN was able to locate and segment the primary tumor on the T2-weighted images in 94% of cases. There was an extremely high agreement between the automatic tool and the manually edited masks. This is the first study to validate an automatic segmentation model for neuroblastic tumor identification and segmentation with body MR images. The semi-automatic approach with minor manual editing of the deep learning segmentation increases the radiologist’s confidence in the solution with a minor workload for the radiologist.

Neurotoxicity is an off-tumour, on-target side effect of GD2-directed immunotherapy with monoclonal antibodies. Here, we report the frequency, management and outcome of patients enrolled in two prospective clinical trials who experienced severe neurotoxicity during immunotherapy with the anti-GD2 antibody dinutuximab beta (DB) administered as short-term infusion (HR-NBL1/SIOPEN study, randomisation R2, EudraCT 2006-001489-17) or as long-term infusion (HR-NBL1/SIOPEN study, randomisation R4, EudraCT 2006-001489-17 and LTI/SIOPEN study, EudraCT 2009-018077-31), either alone or with subcutaneous interleukin-2 (scIL-2). The total number of patients included in this analysis was 1102. Overall, 44/1102 patients (4.0%) experienced Grade 3/4 neurotoxicities (HR-NBL1 R2, 21/406; HR-NBL1 R4, 8/408; LTI study, 15/288), including 27 patients with severe neurotoxicities (2.5%). Events occurred predominantly in patients receiving combined treatment with DB and scIL-2. Neurotoxicity was treated using dexamethasone, prednisolone, intravenous immunoglobulins and, in two patients, plasmapheresis, which was highly effective. While neurological recovery was observed in 16 of 21 patients with severe neurotoxicities, 5/1102 (0.45%) patients experienced persistent and severe neurological deficits. In conclusion, severe neurotoxicity is most commonly observed in patients receiving DB with scIL-2. Considering the lack of clinical benefit for IL-2 in clinical trials so far, the administration of IL-2 alongside DB is not recommended.