RESUMEN
PURPOSE: More than 95% of children with B-lineage acute lymphoblastic leukemia (ALL) achieve a clinical remission after the induction phase of chemotherapy (first 28 days) as evaluated by morphologic criteria. However, relapse occurs in approximately 30% of these children. The objective of this study was to determine whether the outcome of patients in clinical remission at the end of induction therapy could be predicted using a highly sensitive method to detect residual disease. PATIENTS AND METHODS: All children diagnosed with B-lineage ALL at the Children's Hospital of Philadelphia during a 2-year period were eligible. The extent of residual leukemia was quantitated in remission marrow samples obtained at the end of induction therapy in 44 children using a phage clonogenic assay in association with complementarity-determining-region 3 (CDR3)-polymerase chain reaction (PCR). RESULTS: Residual disease was a significant predictor of outcome independent of WBC count, age, or sex. The estimated relapse-free survival (RFS) during therapy was 50.4% (+/- 12.6%) for patients with high residual disease (> or = 0.6% leukemia cells among total marrow B cells) versus 91.9% (+/- 5.5%) for those with lower levels (P < .002). There were no significant differences in off-treatment RFS between patients with high or low residual disease who completed therapy in continuous remission (P = .82). The overall estimated RFS was 32.3% (+/- 11.6%) for patients with high residual disease versus 62.6% (+/- 10.7%) for patients with lower levels of residual leukemia cells, with a median follow-up of 5.3 years for patients in continuous remission (P < .008). CONCLUSION: PCR detection of high residual disease at the end of induction therapy identifies patients at increased risk for relapse during therapy.