First Isolated Extramedullary Relapse in Children with B-Cell Acute Lymphoblastic Leukemia: Results of the Coprall-97 Study.

Isolated extramedullary relapse is diagnosed in 2 to 6% of children with acute lymphoblastic leukemia (ALL). It usually occurs in the CNS or the testis. Because the prophylaxis of extramedullary leukemia has increasingly become an important part of the first line treatment, the ability to achieve a sustained second remission is now complicated by drugs toxicities and resistances. It is already known that prognosis depends on duration of first complete remission (CR1).Here, we report on the efficiency of a combination of chemotherapy and radiotherapy for the treatment of extramedullary relapses. Furthermore, we specifically searched for prognosis factors of the Event Free Survival (EFS). Between May 1997 and December 2002, 68 children and adolescents up to 20 year-old with first isolated extramedullary relapse (group G3) of B-cell (non Burkitt) ALL were included in the prospective, stratified, and multicentric trial COPRALL-97 This trial had been designed for patients pretreated with an intensive frontline therapy (EORTC or FRALLE) without cranial radiotherapy except for 5 patients. Stratification criteria was time to relapse: CR1 of less than 24 months for group G3A (n=35), relapse beyond 24 months for group G3B (n=33). Treatment consisted of risk-adapted alternating short course multiagent chemotherapy through systemic (with good CSF penetration) and intrathecal ways and conventional maintenance therapy and irradiation (18 Gy). Stem cell transplantation (SCT) was performed if an HLA identical related donor was avalaible in group G3A. Sixty four of 68 eligible patients achieved a second complete remission. There were 45 CNS relapses, 22 testis relapses and one ovary relapse. In group G3A (33 CNS, 2 testis), 20 patients received chemotherapy only and 15 had SCT (12 HLA-identical sibling donor, 2 HLA-identical unrelated donor and one autologous. In group G3B, (12 CNS, 20 testis, 1 ovary), all patients but one (SCT not indicated) were treated by chemotherapy. EFS and overall survival (OS) for all registered patients at 8 years were 40% and 53% respectively. EFS at 8 years for patients of group G3A and G3B were respectively 26% and 51% (p=0,0071) while OS at 8 years were respectively 40% and 68% (p=0,065). Our analyses highlighted two independant risks factors predictive of decreased EFS: early relapse and age at the initial diagnosis above 6 years. All other factors studied (sex, site of relapse, immunophenotype, SCT) did not predicted outcome. Most of the second relapses involved bone marrow (n=11), CNS (n=6) or were combined (n=6). There were 6 toxic deaths.We conclude that more than half of the patients who were treated with COPRALL-97 therapy had long term survival. Time of relapse and age at initial diagnosis represent important factors that should be considered when adapting treatment intensity to individuals in future trials. Furthermore, as already reported by others, early CNS relapses have a bad prognosis and new therapeutic strategies are needed.