This single-center retrospective single-arm observational study included patients with high-risk NB who received dinutuximab beta as first-line maintenance therapy between May 2021 and July 2024 in Beijing Children’s Hospital, Capital Medical University. All patients underwent long-term continuous infusion of dinutuximab beta in combination with subcutaneous GM-CSF, and oral isotretinoin, with or without chemotherapy. The primary outcome was the objective response rate (ORR), and the secondary outcomes included 1- and 2-year EFS and OS rates, and safety. The cutoff date was January 2025.

This study was approved by the Institutional Review Board of Beijing Children’s Hospital (2021-E-029-Y) and was conducted in accordance with the principles of the Declaration of Helsinki. Written informed consent was waived.

Eligible patients were aged 1–18 years and diagnosed with high-risk NB (stage 4) according to the International Neuroblastoma Risk Group Staging System (20). All patients received induction therapy (chemotherapy and surgery), followed by consolidation treatment with or without ASCT and/or radiotherapy (Supplementary Tables S1, S2), in accordance with the Beijing Children’s Hospital NB treatment guidelines (BCH-NB-2019) (10). Inclusion required complete response (CR) or partial response (PR) before initiating dinutuximab beta and at least one post-immunotherapy response assessment. Patients with other primary malignancies were excluded.

Patients received five 35-d cycles of dinutuximab beta (100 mg/m² per cycle) administered as a continuous intravenous infusion at a dose of 10 mg/m²/d. Two regimens have been used in clinical practice: (1) For patients achieving CR after prior therapies, dinutuximab beta in combination with subcutaneous GM-CSF and oral isotretinoin were administered (regimen 1). (2) For patients achieving PR after prior therapies or those in CR who did not undergo ASCT, dinutuximab beta with GM-CSF, oral isotretinoin, and chemotherapy [either irinotecan/temozolomide (IT) or topotecan/temozolomide (TT)] were administered (regimen 2). In regimen 1, dinutuximab beta was administered on days 4–13, subcutaneous GM−CSF 250 μg/m²/d on days 1–14, and oral isotretinoin (160 mg/m²/d, divided into two daily doses) on days 16–29. In regimen 2, dinutuximab beta was administered on days 1–10, alongside GM-CSF at 250 ug/m²/d subcutaneously on days 6–12, and oral isotretinoin (160 mg/m²/d, divided into two daily doses) on days 15–28. The IT regimen comprised irinotecan (50 mg/m²/d, intravenously) and temozolomide (100 mg/m²/d, orally) on days 1–5; for patients intolerant to IT, the TT regimen substituted topotecan (0.75 mg/m²/d, intravenously, days 1–5) with temozolomide dosed as above. Considering the tolerability of combination therapy, dinutuximab beta was not combined with chemotherapy in the first cycle, but was initiated in the second chemotherapy cycle.

Gabapentin and morphine were administered to manage pain. During the first cycle of dinutuximab beta treatment, all patients received intravenous morphine via an infusion pump to prevent and control pain. Morphine dosing was individualized and gradually tapered, transitioning from intravenous to oral administration and subsequent discontinuation.

Glucocorticoids, intravenous immunoglobulins, and investigational agents were not administered during dinutuximab beta treatment. In cases of disease progression or relapse during therapy or follow-up, patients could receive additional antitumor therapy such as chemotherapy or local radiotherapy.

Tumor response was assessed according to the 2017 International Neuroblastoma Response Criteria (21). Efficacy evaluations were performed at baseline (2–4 weeks before immunotherapy treatment initiation), after the 2nd or 3rd cycle, at the end of treatment (EOT), and every 3 months thereafter for up to 3 years. Imaging assessments included meta-iodobenzylguanidine (MIBG) scintigraphy, positron emission tomography/computed tomography (PET/CT), ultrasound, contrast-enhanced cranial magnetic resonance imaging (MRI) or computed tomography (CT) of both primary and metastatic lesions. Bone marrow (BM) was assessed using standard morphological cytology, PHOX2B molecular testing, and flow cytometry. MIBG scintigraphy was evaluated using the Curie score, which semi-quantifies skeletal disease extent within nine body segments on a 0 to 3 scale.

Safety assessments included the documentation of adverse events (AEs), pain severity, vital signs, and laboratory parameters. AEs were graded according to the Common Terminology Criteria for Adverse Events, version 5.0. Pain was measured using age-appropriate scales: the Wong-Baker Faces Pain Rating Scale for patients aged 3–8 years and the Face, Legs, Activity, Cry, Consolability scale for patients younger than 3 years. Opioid dose and duration (intravenous or oral morphine) were also recorded.

The ORR was defined as the proportion of patients achieving CR or PR. ORR also were assessed at the subgroup level. EFS was defined as the time from the initiation of dinutuximab beta to progressive disease, relapse, secondary tumor, or death from any cause, whichever event occurred first. OS was defined as the time from the initiation of dinutuximab beta to death from any cause. The 1- and 2-year survival rates [with 95% confidence interval (CI)] were estimated using the Kaplan–Meier method, with group comparisons using the log-rank test. Survival analyses were stratified according to prior response status (CR/PR), MYCN amplification (yes/no), underwent ASCT (yes/no), and treatment regimens (regimen 1/regimen 2).

Descriptive statistics were used to summarize the baseline characteristics, treatments, and safety. Continuous variables were reported as mean, standard deviation, median, and interquartile range; categorical variables were reported as frequencies and percentages. All analyses were performed in SPSS v22.0, with supplementary checks in R where indicated. Two-sided p < 0.05 was considered nominally significant.

Twenty-eight eligible patients received dinutuximab beta, among them, 14 had achieved CR and 14 PR prior to immunotherapy (Figure 1). All three CR patients with prior ASCT received regimen 1 (without chemotherapy). Of the 11 CR patients without ASCT, eight followed the scheduled regimen 2 (with chemotherapy), while three received regimen 1 instead. For the 14 PR patients, regimen 2 was the protocol-specified treatment; however, four patients (one with ASCT and three without ASCT) received regimen 1. Overall, seven patients (three CR without ASCT and four PR) deviated from the treatment protocol. Because three were ineligible for chemotherapy due to poor organ function, and chemotherapy was also avoided in the other four patients - the first cases in the newly introduced dinutuximab beta in China - to minimize risk amid limited clinical experience.

Table 1 shows that 17 patients (60.7%, 17/28) were male, the median age at diagnosis was 38 months (range 17–101), and that at the initiation of dinutuximab beta was 56 months (range 27–113). MYCN amplification occurred in 10 patients (35.7%). Only six patients (21.4%) underwent ASCT.

At the end of induction therapy, 13 patients (46.4%) achieved CR and 15 (53.6%) achieved PR. One patient with a minimal residual lesion in the right temporal bone (initially evaluated as PR) achieved lesion resolution after radiotherapy and chemotherapy. Before dinutuximab beta initiation, 14 patients (50.0%) had CR and 14 (50.0%) had residual disease (PR). Among the latter, seven (50.0%) had bone-only metastases, four (28.6%) had both bone and soft tissue metastases, two (14.3%) had bone and BM residual disease, and one (7.1%) had soft tissue metastases. Dinutuximab beta was combined with chemotherapy in 18 patients (64.3%), including 8 of CR and 10 of PR.

Compared with patients in CR prior to immunotherapy, those in PR were older at both diagnosis (29.5 vs. 43.5 months) and the initiation of dinutuximab beta (40.5 vs. 59.5 months). MYCN amplification was more frequent in patients with CR prior to dinutuximab beta immunotherapy than in those with PR (57.1% vs. 14.3%).

Among the patients with CR prior to immunotherapy, 11 (78.6%, 11/14) maintained CR at the last follow-up. The remaining three patients experienced one death from intracranial metastases, one sphenoid bone relapse, and one relapse in the mediastinal and cervical lymph nodes (Supplementary Table S1).

Among the 14 patients with residual disease prior to immunotherapy, the ORR was 64.3% (9/14) at EOT and 57.1% (8/14) at the 1-year follow-up (Table 2). Stratified by treatment modalities, patients who received dinutuximab beta with chemotherapy had a higher ORR than those who received dinutuximab beta alone [EOT: 70.0% (7/10) vs. 50.0% (2/4); 1-year follow-up: 60.0% (6/10) vs. 50.0% (2/4)]. Based on ASCT status, 11 patients without ASCT had an ORR of 54.5% (6/11), whereas all three patients with ASCT achieved CR at EOT.

Among the 13 patients with bone involvement, the ORR was 61.5% (8/13) at EOT and 53.9% (7/13) at 1-year follow-up (Table 3). Of the five patients with soft tissue involvement, three achieved CR at EOT and maintained CR until the last follow-up, resulting in an ORR of 60.0% (3/5) at both EOT and 1-year follow-up. Notably, both patients with minimal residual BM disease achieved CR (100%) at cycle 2/3 and maintained CR through the last follow-up.

Among patients with residual disease who underwent MIBG scans (n =12), the mean baseline and EOT MIBG scores were 2.4 (range 1–9) and 0.5 (range 0–2), respectively with a mean reduction of –1.9 (range –9 to 0). Individual MIBG scores are shown in Figure 2.

Lesion response was achieved by 66.7% of the patients (8/12) at EOT, who remained disease-free at the 1-year follow-up (Patients 15, 16, 17, 18, 19, 25, 27, and 28, Supplementary Table S3). Patient No. 15 achieved CR at EOT and 1-year follow-up but subsequently relapsed. Four patients had persistent lesions throughout the final follow-up period (Patients 20, 21, 23, and 26).

At a median follow-up of 19 months (range 6–44), the estimated 1- and 2-year EFS rates for the entire cohort were 88.2% (95% CI, 67.4–96.1%) and 72.4% (95% CI, 41.4–88.8%), respectively; the 1- and 2-year OS rates were both 95.8% (95% CI, 73.9–99.4%) (Figures 3A, B).

In the subgroup analysis, patients in prior CR (CR subgroup) showed a lower 1-year EFS rate than those in the PR subgroup [85.7% (95% CI, 53.9–96.2%) vs. 90.9% (95% CI, 50.8–98.7%), p = 0.684], but a higher 2-year EFS rate [76.2% (95% CI, 42.1–91.8%) vs. 68.2% (95% CI, 16.3–92.2%), p = 0.747]. One- and 2-year OS rates were lower in the CR subgroup than in the PR subgroup [91.7% (95% CI, 53.9–98.8%) vs. 100%, p = 0.307; Figures 3C, D].

Table 4 shows that patients with MYCN amplification had lower 2-year EFS rate [70.0% (95% CI, 22.5–91.8%) vs. 80.8% (95% CI, 50.5–93.6%), p = 0.620] but numerically higher 2-year OS rate [100% vs. 93.3% (95% CI, 61.3–99.0%), p = 0.437] than those without amplification. Patients who underwent ASCT showed numerically better 2-year EFS and OS rates than those who did not [EFS: 83.3% (95% CI 27.3–97.5%) vs. 68.8% (95% CI, 31.6–88.6%), p = 0.498; OS: 100% vs. 94.7% (95% CI, 68.1–99.2%), p = 0.602]. Patients receiving dinutuximab beta with chemotherapy were inclined to achieved better 2-year EFS [90.9% (95% CI, 50.8–98.7%) vs. 51.4% (95% CI, 14.3–79.6%), p = 0.130] and OS rates [100% vs. 88.9% (95% CI, 43.3–98.4%), p = 0.185] compared to those receiving dinutuximab beta alone. Additionally, patients treated with DFMO following dinutuximab beta had numerically lower 2-year EFS rate [67.7% (95% CI, 41.2-100.0%) vs. 81.7 (95% CI, 60.8-100.0%), p = 0.524] but higher 2-year OS rate [100.0% vs. 90.9% (95% CI, 75.4-100.0%), p = 0.317] than those who did not receive DFMO.

All patients experienced AEs, which were most frequent and severe during the first treatment cycle and decreased in subsequent cycles (Figures 4, 5). Most AEs were of grade 1 or 2, and grade 3 or 4 events occurred predominantly in patients who received combination chemotherapy.

In patients receiving dinutuximab beta without chemotherapy (n = 10), common grade 1 or 2 AEs included low albumin (90%), fever (80%), diarrhea (50%), chills (50%), and elevated g-glutamyl transferase (GGT, 50%, Figure 4) in cycle 1. Grade 3 events included fever (20%), elevated GGT levels (20%), infection (10%), anemia (10%), thrombocytopenia (10%), and elevated alanine aminotransferase (10%). Fever, eye abnormalities, and thrombocytopenia persisted across five cycles (Supplementary Table S4).

Patients receiving dinutuximab beta with chemotherapy (n = 18) experienced more frequent and severe AEs, particularly hematological toxicities (Figure 5). In the first administration cycle combined with chemotherapy (cycle 2), the common grade 1 or 2 AEs included fever (72.2%), low albumin (83.3%), anemia (66.7%), and neutropenia (66.7%). Grade 3 neutropenia occurred in 22% of patients and persisted in later cycles. Occasional grade 4 events including fever, neutropenia, and thrombocytopenia were reported. Non-hematological AEs were mostly mild to moderate and included pain, diarrhea, and oral mucositis (Supplementary Table S5). Despite the higher AE burden associated with combination chemotherapy, all events were manageable with supportive care, and no treatment-related deaths occurred.

Pain was the most common dinutuximab beta-related AE. The median pain scores in cycle 1 and mean daily morphine doses across all cycles are shown in Figure 6. Intravenous morphine peaked at 0.75 mg/kg/d on day 3 of cycle 1, corresponding to the highest median pain score of 5. By day 4, half of the patients switched to oral morphine, which tapered to 0.05 mg/kg/d by day 8. Morphine requirements in cycles 3–5 were minimal, with low oral doses of 0.01–0.05 mg/kg/d. Pain was effectively controlled with a combination of gabapentin and morphine, allowing all patients to complete the treatment without discontinuation owing to pain.