Antibody supply was discontinued by United Therapeutics Corporation.

Although the five year survival rate of children with high risk neuroblastoma have increased over the last three decades from 4 to 44 % (1), neuroblastoma is the second most frequent cause for cancer related death in childhood (11 %). Most patients show good initial response rates (complete (CR) and partial remission (PR) rate 95 %), but 55 % experience a largely treatment-resistant tumor progression.

Recently, a breakthrough with immunotherapy was reported by US investigators from the Children's Oncology Group (2) using the anti-ganglioside D2 (GD2) monoclonal antibody ch14.18 for tumor cell destruction and granulocyte macrophage-colony stimulating factor (GM-CSF) plus interleukin 2 (IL-2) for immunostimulation. This immune therapy resulted in an increase of 20 % Event free survival (EFS) at 2 year from randomization. However, this was associated with a high toxicity rate (pain, capillary leak syndrome).

The proposed trial compares the Childrens' Oncology Group (COG) "standard of care" arm (anti-GD2 + GM-CSF + IL-2 i.v. + retinoic acid oral) with an experimental arm (anti-GD2 + GM-CSF + IL-2 s.c. + retinoic acid oral) designed to reduce toxicity.

The potential benefit from this trial consists of the confirmation that the American trial design is feasible in an independent set of patients with different preceding therapy, at a different time point regarding to immune reconstitution after autologous stem cell transplantation (ASCT), the feasibility of a newly designed immunotherapy (which is hopefully less toxic) and the investigation of immune response parameters. This pilot study is the prerequisite for a consecutive randomized clinical trial comparing two immunotherapeutic approaches in a larger set of patients.

Neuroblastoma is the second most frequent solid tumour (7.6%) and the second cause of cancer related death (11%) in childhood. In particular the large high risk (HR) group has remained a challenge to Paediatric Oncologists. Although the 5 year survival rates of children with HR disease have increased over the last 3 decades from 4 to 44.4%, the vast majority of those children will finally succumb to disease [1]. Most patients show good initial responses to chemotherapy (CR + PR-rate 95 %), but a majority experiences a highly treatment resistant tumour progression (55%). Therefore new therapeutic modalities are urgently needed.

Recently, a randomized trial demonstrated that an immunotherapeutic concept using the anti-GD2 antibody ch14.18 together with interleukin 2, GM-CSF and retinoic acid improved the outcome of neuroblastoma patients which achieved CR or very good partial remission (VGPR) response to the preceding therapy. This treatment was associated with a high rate of toxic effects (neuropathic pain 52% of patients, capillary leak syndrome 23%). An earlier study using the antibody ch14.18 alone made comparable observations: pain despite of analgesia was seen in 33% of patients and severe capillary leak syndromes in 3 of 151 children. Differences of the reported frequencies are due to the different definitions of side effects.

The investigators therefore propose a randomized clinical trial comparing the COG immunotherapy concept with newly designed hopefully equally effective but less toxic concept. This modifies the application route of IL-2 from i.v. to s.c. and increases the IL-2 dose to 6.0 mio/m²xd (from 3.0 first week and 4.5 mio second week i.v.). Oral retinoic acid is used in both arms. The proposed randomized trial will answer the following questions:

(i) Confirmation of the feasibility to apply the COG immunotherapy as consolidation treatment after a different remission induction therapy in patients with recurrent and de novo high risk neuroblastoma (ii) Investigation of the feasibility to apply the new immunotherapy concept in patients with recurrent high risk neuroblastoma (iii) Comparison of the toxicity of both immunotherapy regimens with the aim to reduce grade 2 - 4 toxicities in the experimental arm.

(iv) Comparison of the immune response (antiidiotype antibodies, immune cell phenotypes, immune mediators, functional assays as antibody dependent cellular cytotoxicity (ADCC) and complement dependent cytotoxicity (CDC) between treatment cycles (intraindividual), treatment arms (interindividual) and between recurrent and newly diagnosed patients (v) Comparison of pharmacokinetics of antibody ch14.18 in both arms (12.10.) (vi) Comparison of therapeutic efficacy by response evaluation at the end of the 25 week treatment (descriptive).

(vii) Comparison of patients´ QoL experienced in both immunotherapy regimens as indicated by parents rating in appropriate questionnaires.

Inclusion Criteria:

• Established diagnosis of neuroblastoma according to the international criteria (INSS)
• High risk (HR): stage 4 over 18 months of age and oncogen MYCN (MYCN) amplified neuroblastoma of any stage and any age until 25 years (recurrent disease (Germany and The Netherlands) after re-induction chemotherapy (+/- other modalities) or newly diagnosed disease (only The Netherlands):
• Complete front-line treatment including induction chemotherapy, radioisotope (mIBG) treatment, appropriate local therapy such as surgical removal and/ or local irradiation of the primary tumor and myeloablative chemotherapy with autologous stem cell reinfusion according to the actual guidelines of the GPOH/DCOG
• achieved response status: stable disease or better (CR, VGPR, PR, SD).
• Written informed consent of parents or guardian and - if appropriate - of the patient.
• For at least two weeks prior to start of trial medication off any standard or experimental treatment no tumour surgery no immediate requirements for palliative chemotherapy, radiotherapy or surgery
• The patient may have had prior central nervous system (CNS) metastases provided the following criteria are all met:

The patient's CNS disease has been previously treated The patient's CNS disease has been clinically stable for four weeks prior to starting this study (assessed clinically and by MRI or CT) The patient is off steroids for four weeks prior to starting the study and will not require them during the course of the study A patient with seizure disorders may be enrolled if well controlled on anticonvulsants and if no seizures have occurred within a 6 week period prior to starting trial treatment

• HIV sero-negative and neither active nor chronic-replicative hepatitis B infection
• Laboratory testing: The patients should have adequate functions of the cor, lung, bone marrow, liver, kidney

Exclusion Criteria:

-