Prognostic Factors for Relapsed Childhood Acute Lymphoblastic Leukemia (ALL): A Report from Tokyo Children’s Cancer Study Group (TCCSG) Study L95-14 and L-99-15.

More than 70% of children with ALL enjoy a long-term event-free survival (EFS), however, the rest of them still suffer from a relapse and the treatment for relapsed ALL is not always successful. The treatment results of childhood ALL with a first relapse were retrospectively analyzed to determine prognostic factors.

Of the 1336 children enrolled on the TCCSG L95-14 and L99-15 studies, diagnosed between 1995 and 2003, 224 suffered from a relapse before March 2005. The relapsed patients were treated differently according to the protocol of each institution. The potential prognostic factors examined were : the age (low; 1–9 years, high; over 10 years), white blood cell count (low; <50,000/μl, high; others) and immunophenotype of leukemic blasts at initial diagnosis (B-lineage or T), gender, initial risk groups, the site of relapse, the timing of relapse (very early: within 18 months after diagnosis, early: other than either very early or late relapse, late: later than 6 months after the discontinuation of the front-line chemotherapy), and the treatment after relapse (chemotherapy alone or chemotherapy plus SCT).

The sites of relapse were: isolated bone marrow (BM, n=165), BM with other sites (n=21) and extramedullary only (n=38). Ninety-nine relapses occurred in very early stage, 43 in early stage and 82 in late stage. Twenty-three patients received SCT before relapse according to the protocol.

A second complete remission (2CR) was achieved in 175/220 patients (79%) and 74 of them are still in 2CR. Five-year EFS for all the relapsers was of 29±3% (mean±s.e.) and 5-year overall survival rate was 38±4%. In a subset of patients (n=126) with B-lineage phenotypes and with isolated BM relapse, who were treated with chemotherapy alone before relapse, we found 2 significant prognostic factors identified by a multivariate analysis : the age at initial diagnosis (EFS: age low; 35±5%, high; 8±8%, p=0.0016) and the timing of relapse (EFS: late; 39±8%, early; 24±9%, very early; 16±5%, p=0.0051). Gender, initial leukocyte count, or the use of SCT after obtaining 2CR was not related with patient’s outcome in a multivariate analysis.

The importance of the timing of relapse was overwhelmingly shown, however, the significance of the age at the diagnosis was not demonstrated before. The results will be incorporated into our next protocols to relapsed patients with ALL.