BACKGROUND: The National Chilean Pediatric Oncology Group, PINDA, reports the first prospective, nonrandomized trial for acute lymphoblastic leukemia (ALL), using a modified version of the Berlin-Frankfurt-Munster protocol (ALL BFM 86). The aim of this study was to classify immunophenotypes, to decrease cranial irradiation, and to assess whether this protocol would improve the survival rate. PROCEDURE: From June, 1987, to June, 1992, 444 unselected children were diagnosed with ALL. Of them, 425 were evaluable. Therapy was stratified by risk. Standard-risk (SR) and high-risk (HR) patients received protocols I, M, II, and maintenance therapy. Very-high-risk (VHR) patients received protocol E instead of protocol M. All patients received a prephase treatment consisting of prednisone and intrathecal methotrexate (MTX). HR and VHR patients received cranial irradiation (12-18 Gy). The following changes were made to the ALL BFM 86 protocol: in protocol M, MTX 1 g/m2 instead of 5 g/m2; in protocol E, citarabine 1 g/m2 instead of 2 g/m2; mithoxantrone and ifosfamide were substituted by teniposide and cyclophosphamide. RESULTS: Immunophenotypes: pro-B-ALL, 14%; common ALL, 67.4%; pre-B-ALL, 4.3%; T-ALL, 10%; undifferentiated leukemia (AUL), 4.3%. The overall 5-year event-free survival (EFS) rate was 60% +/- 2% (SE). The 5-year EFS rate for each risk group was: SR 75%, HR 62%, VHR 28%, with a median follow-up of 6.5 years (range 4.5-9.5 years). The cumulative incidence of central nervous system (CNS) relapse was 5.4%. CONCLUSIONS: We have been able successfully to perform a nationwide study. Our strategy to adapt the BFM protocol to our population of patients trial was effective in improving the EFS. The immunophenotype distribution is similar to that in other reported series.