RATIONALE: Drugs used in chemotherapy work in different ways to stop the growth of tumor cells, either by killing the cells or by stopping them from dividing. Giving more than one drug (combination chemotherapy) may kill more tumor cells. Giving combination chemotherapy before surgery may make the tumor smaller and reduce the amount of normal tissue that needs to be removed. Sometimes, after surgery, the tumor may not need more treatment until it progresses. In this case, observation may be sufficient. It is not yet known which combination chemotherapy regimen given together with surgery, with or without autologous bone marrow or peripheral stem cell transplant, is more effective in treating newly diagnosed neuroblastoma.

PURPOSE: This phase III trial is studying combination chemotherapy to see which regimen given together with surgery, with or without autologous bone marrow or peripheral stem cell transplant, works in treating infants with newly diagnosed neuroblastoma.

OBJECTIVES:

Primary

• Determine the outcome, in terms of survival and morbidity, in infants with localized, unresectable, non-MYCN-amplified neuroblastoma treated with reduced-intensity chemotherapy.
• Determine the survival of infants with stage 4S neuroblastoma, no MYCN amplification, and no bone, CNS, or pleural/lung metastases treated with short-course intensive chemotherapy.
• Determine the survival of infants with stage 4S neuroblastoma, no MYCN amplification, and bone, CNS, or pleural/lung metastases not treated with intensive high-dose chemotherapy consolidation.
• Determine the survival of infants with any stage (except stage 1) neuroblastoma and MYCN amplification treated with intensive consolidation high-dose chemotherapy followed by autologous bone marrow or stem cell support.

Secondary

• Correlate outcome with factors other than stage and MYCN status in infants with neuroblastoma.
• Define the behavior of neuroblastoma in infants treated with these regimens.
• Determine prognostic criteria in infants treated with these regimens.
• Determine whether deletion of chromosome 1p or diploidy/tetraploidy are prognostic factors in infants who do not have other adverse features, such as MYCN amplification.

OUTLINE: This is a nonrandomized, multicenter study. Patients are assigned to 1 of 4 treatment regimens according to disease criteria. Patients who are not eligible for any of these regimens (stage 1 or resectable stage 2 disease) undergo surgical resection followed by observation.

• Regimen NB 99.1 (unresectable stage 2 or 3): Patients are treated according to spinal cord involvement and presence of neurological symptoms.

  • Group I (no evidence of spinal cord involvement):

    • CO therapy: Patients receive cyclophosphamide IV on days 1-5 and vincristine IV on day 1. Treatment repeats every 14 days for up to 4 courses in the absence of disease progression. Resectability is assessed after every 2 courses of chemotherapy; if tumor is resectable, then patients undergo surgery followed by observation only. If, after 4 courses of CO, the tumor is still not resectable or disease has progressed, then patients proceed to VP-CARBO therapy.
    • VP-CARBO therapy: Patients receive carboplatin IV over 1 hour and etoposide phosphate IV over 2 hours on days 1-3. Treatment repeats every 21 days for 2 courses. If the tumor is then deemed resectable, the patient undergoes surgery. If the tumor is not resectable or disease has progressed, then patients proceed to CADO therapy.
    • CADO therapy: Patients receive cyclophosphamide IV over 1 hour on days 1-5, doxorubicin hydrochloride IV over 6 hours on days 4 and 5, and vincristine IV on days 1 and 5. Treatment repeats every 21 days for 2 courses. Patients then proceed to resection or biopsy.

  • Group II (dumbbell tumors, spinal cord compression symptoms or life-threatening symptoms [e.g., respiratory obstruction]): Patients receive 2 courses of VP-CARBO therapy. Patients who achieve a response proceed to surgery or biopsy if the extraspinal portion is resectable. Patients with nonresponding disease or an unresectable extraspinal portion of the tumor receive 2 courses of CADO therapy and then undergo surgery or biopsy. Patients with dumbbell tumors but no spinal cord compression symptoms are treated as in group I.

• Regimen NB 99.2 (stage 4S or stage 4 without bone, pleura/lung, or CNS metastases and no MYCN amplification): Patients who do not have severe or life-threatening symptoms are observed for spontaneous regression of disease. Patients with severe symptoms receive 1 course of VP-CARBO therapy. Patients with a Philadelphia score ≥ 2 (or ≥ 1 for neonates [< 1 month old]) receive a second course of VP-CARBO therapy. If disease does not respond to 2 courses of VP-CARBO therapy, patients receive up to 4 courses of CADO therapy. Treatment ceases after response is obtained. Surgery is allowed but not required.

• Regimen NB 99.3 (skeletal bone, pleural, and/or CNS metastases, no MYCN amplification): Patients receive 2 courses of VP-CARBO therapy. Patients with responding disease receive 2 more courses and then proceed to surgery (if possible). Patients with disease progression or no response after the first 2 courses of VP-CARBO therapy and patients who do not experience metastatic complete response (CR) after 4 courses of VP-CARBO therapy receive up to 4 courses of CADO therapy. Patients proceed to surgery, if possible, after 2-4 courses of CADO therapy.

• Regimen NB 99.4 (stages 2-4 disease with MYCN amplification): Patients receive 2 courses of VP-CARBO therapy followed by 2 courses of CADO therapy and then surgery (if not already performed). Patients receive filgrastim (G-CSF) subcutaneously for 5 days between the second course of CADO therapy and surgery. Patients also undergo collection of their bone marrow or peripheral blood stem cells (PBSC). Patients who undergo surgery receive 1 course of VP-CARBO therapy followed by 1 course of CADO therapy postsurgery. At least 3 weeks after the third course of CADO therapy, patients receive high-dose chemotherapy comprising busulfan every 6 hours on days -7 to -3 and melphalan IV on day -2 followed by autologous bone marrow or PBSC infusion on day 0. At least 2 months later, patients undergo radiotherapy to the primary tumor site, even if complete surgical resection has been accomplished. Patients with stage 4 disease who do not achieve metastatic CR after chemotherapy (before surgery) go off study.

After completion of study treatment, patients are followed periodically.

PROJECTED ACCRUAL: A total of 330 patients will be accrued for this study.

DISEASE CHARACTERISTICS:

• Histologically confirmed neuroblastoma or ganglioblastoma

  • Newly diagnosed disease

• MYCN status known

• Patients are assigned to different study treatment regimens* according to the following disease criteria:

  • Regimen NB 99.1

    • Localized unresectable tumor

      • If open surgery is considered too hazardous due to the site of the tumor or condition of the infant, then cytological confirmation is allowed, provided adequate material is obtained for study procedures

    • No MYCN amplification (i.e., < 10 copies)

    • No metastatic deposits in the bone marrow

    • No MIBG or technetium uptake or radiological bone lesions in the skeleton

    • No liver disease by ultrasound

  • Regimen NB 99.2

    • Stage 4 or 4S with metastases confined to the skin, marrow, nodes, or liver

      • No bone involvement by radiographs
      • No pleural or lung involvement
      • No CNS involvement

    • No MYCN amplification (i.e., < 10 copies)

  • Regimen NB 99.3

    • Stage 4 disease, metastases must meet ≥ 1 of the following criteria:

      • Skeletal bone metastases by plain x-ray or CT scan
      • Pleural or lung metastases
      • CNS involvement

    • No MYCN amplification (i.e., < 10 copies)

  • Regimen NB 99.4

    • Stage 2-4 disease
    • MYCN amplification (i.e., ≥ 10 copies) NOTE: *Patients with stage 1 or resectable stage 2 disease with no MYCN amplification ( i.e., < 10 copies) are not eligible for a study treatment regimen but may be enrolled on the study for observation only

PATIENT CHARACTERISTICS:

• Not specified

PRIOR CONCURRENT THERAPY:

• No prior therapy