Long-term outcome and clinical prognostic factors in children with medulloblastoma treated in the prospective randomised multicentre trial HIT'91

To analyse long-term outcome and clinical prognostic factors in medulloblastoma. We analysed 280 patients with medulloblastoma (3-18 years) included from 1991 to 1997 in the randomised multicentre trial HIT'91 comparing pre-('sandwich') and postradiation ('maintenance') chemotherapy (median follow-up of survivors for 10 years). In 187 patients with complete staging, overall survival (OS) was higher after maintenance compared to sandwich treatment for M0 (10-year OS 91% and 62%, p=0.001) and M1 patients (10-year OS 70% and 34%, p=0.020). In M2/3 disease, 10-year OS was 42% and 45%. Incomplete staging, metastases, younger age and sandwich chemotherapy were independent adverse risk factors. Twelve percent of all relapses (13 of 107) occurred after more than five years, and 12 patients had secondary neoplasms. After maintenance therapy, long-term survival was excellent in fully assessable patients with localised medulloblastoma, and favourable for M1 patients. Patients should be followed longer for late relapses and secondary tumours.