Minimal residual disease tests provide an independent predictor of clinical outcome in adult acute lymphoblastic leukemia

Purpose: Investigation of minimal residual disease (MRD) in childhood acute lymphoblastic leukemia (ALL) using molecular markers has proven superior to other standard criteria (age, sex, and WBC) in distinguishing patients at high, intermediate, and low risk of relapse. The aim of our study was to determine whether MRD investigation is valuable in predicting outcome in Philadelphia-negative adult patients with ALL. Patients and Methods: MRD was assessed in 85 adult patients with B-lineage ALL by semiquantitative immunoglobulin H gene analysis on bone marrow samples collected during four time bands in the first 24 months of treatment. Fifty patients received chemotherapy only and 35 patients received allogeneic (n = 19) or autologous (n = 16) bone marrow transplantation (BMT) in first clinical remission. The relationship between MRD status and clinical outcome was investigated and compared with age, sex, immunophenotype, and presenting WBC count. Results: Fisher's exact test established a statistically significant concordance between MRD results and clinical outcome at all times. Disease-free survival (DFS) rates for MRD-positive and -negative patients and log-rank testing established that MRD positivity was associated with increased relapse rates at all times (P < .05) but was most significant at 3 to 5 months after induction and beyond. MRD status after allogeneic BMT rather than before was found to be an important predictor of outcome in 19 adult patients with ALL tested. In patients receiving autologous BMT (n = 16), the MRD status before BOAT was more significant (P = .005). Conclusion: The association of MRD test results and DFS was independent of and greater than other standard predictors of outcome and is therefore important in determining treatment for individual patients. (C) 2002 by American Society of Clinical Oncology.