Despite advancing cure rates in childhood acute lymphoblastic leukemia (ALL), 25-30% of patients eventually succumb to their disease. The outlook is far worse in acute myeloid leukemia (AML): only about 30% of children are cured in chemotherapy programs, and the option of bone marrow transplantation is either not available (because of the lack of suitable donors) or of limited value (because of high rates of marrow relapse, likely due to residual leukemia in the autografts). Early detection of low levels of residual leukemia would benefit both groups of patients. It would permit more timely investigation of retrieval therapy, possibly before the development of multi-drug resistance, as seen in patients with clinically overt relapse, and it would allow the source of relapse to be determined in patients undergoing autologous bone marrow transplantation, whether residual disease in the marrow implant or resistant leukemia remaining after ablative treatment. The hypothesis underlying this proposal is that many cases of acute leukemia possess distinctive combinations of immunologic markers that can be used to detect minimal residual disease (MRD) and thus identify patients requiring alternative treatment. The assay system to be used is capable of detecting 1 leukemic cells among 10,000 normal cells and therefore can quantify leukemia at the 10-4 level (standard morphologic tests for relapse require 1-5% identifiable blasts). In the first specific aim, marrow samples taken at five different intervals from patients with T- ALL, enrolled in a phase III clinical trial, will be assessed for the leukemia-specific TdT/CD3 phenotype. Similar studies will be performed on specimens from B-lineage ALL patients to detect CD19/CD13, CD19/CD33, CD19/CDw65 and CD34/CD21 phenotypes. Comparison of outcome data for children with and without MRD will demonstrate with 80% power the value of this approach for prospective identification of high-risk patients. The large patient accrual (> 200 subjects over 4 years) will also make it possible to directly compare the sensitivity and specificity of our immunologic assays with those of the polymerase chain reaction (PCR), as applied to clonal molecular abnormalities (specific aim 2). PCR analysis has been advocated as a clinically useful tool for MRD detection, but its efficacy relative to that of immunologic marker detection has not been assessed in a controlled, prospective trial. Finally, the usefulness of immunologic methods for MRD detection will be estimated in childhood AML, using marrow specimens from patients scheduled to undergo autologous transplantation (specific aim 3).
Showing the most recent 10 out of 71 publications
