Purpose: NB87 was designed to test the efficacy of a short, non cross-resistant, induction protocol for unselected patients over 1 year of age with stage 4 neuroblastoma. A secondary objective was to compare in a randomized study the toxicity of two modalities of cisplatin administration.
Patients and methods: A total of 183 patients received two cycles of alternating sequences: cyclophosphamide 300 mg/m2/d on days 1 to 5, vincristine 1.5 mg/m2/d on days 1 and 5, and doxorubicin 60 mg/m2/d on day 5 (CADO); and cisplatin 40 mg/m2/d and etoposide 100 mg/m2/d on days 1 to 5 (CVP), followed by surgery of the primary tumor (126 patients). Ninety-one were randomized to receive cisplatin either as bolus (BO; n = 48) or continuous infusion (CI; n = 43). International Neuroblastoma Staging System (INSS) and Response Criteria (INRC) were used with emphasis on skeletal evaluation by meta-iodobenzylguanidine (MIBG).
Results: Hematotoxicity was predominant, with a higher incidence of neutropenia (P = .01) for CADO and of thrombocytopenia for CVP (P < .001). Severe infections, as well as nonhematologic toxicities, occurred more often after the first sequence. Gastrointestinal complications were predominant during both courses of CVP. The toxic death rate, including surgery, was 3%. Complete remissions (CRs) were less frequent on MIBG (45%) compared with marrow (66%) or other metastases (61%). Combining all metastatic sites resulted in a 39% CR rate. After surgery, the final CR rate was 42%. Nephrotoxicity was minimal in both arms (92% normal clearance for CI v 82% for BO). Hearing loss greater than 40 dB at 6,000 to 8,000 Hz was reported equally in both arms (n = 6 for CI v n = 5 for BO).
Conclusion: Intensified chemotherapy using CADO/CVP increases CR rates despite a shorter induction duration. However, the rate of MIBG normalization remains unsatisfactory and could be raised through the dose-intensive use of agents such as cyclophosphamide.