Treatment results in childhood acute lymphoblastic leukemia with a modified ALL-BFM'90 protocol: lack of improvement in high-risk group.

We report results achieved in our institution with an acute lymphoblastic leukaemia risk-oriented treatment trial opened in January 1990 and closed on December 1995. The study was similar to the German ALL-BFM'90, except for using Protocol III for the standard-risk group, 2 g/m2 of methotrexate in Protocol M, and preventive cranial irradiation for the high-risk group only. The high-risk group included mostly patients with prednisone poor initial response and/or adverse cytogenetic features. This analysis included 374 patients, whose mean age was 6 years (range: 1 month-17 years). The overall complete remission rate was 94.4% (353/374) and the 5-year event-free survival (standard error) probability is 64(5)%. The 5-year event-free survival estimates for each risk group were: (1) high-risk group 37(5)%; (2) intermediate-risk group 66(1)%; and (3) standard-risk group 74(4)% (P = 0.0001). There are significantly higher-rates of isolated bone marrow and testicular relapses in the high-risk subset of patients. Our dismal results and the published experience, lead us to conclude that the optimal treatment for these high-risk acute lymphoblastic leukaemia patients is not currently known.