Analysis of minimal residual disease (MRD) in children with lymphoblastic leukemia (ALL): A tool for therapy stratification within the BFM ALL 99 protocol

Background: Children with acute lymphoblastic leukemia are cured in about 70% with current treatment protocols. Although almost all children (>95%) achieve complete remission after induction chemotherapy as defined by cytomorphological criteria, 30% of these children will suffer a relapse. Only few of them are identified at diagnosis as having a high risk for relapse. The in vivo response to chemotherapy as measured by the detection of minimal residual disease (MRD = presence of leukemic cells below the detection limit of morphological methods, i.e. 5%) has been assumed to be of prognostic significance. We were able to show in a prospective multicenter study including a large number of unselected children with ALL, who were treated according to common BFM protocols, that MRD is an independent prognostic parameter. We identified three risk groups according to the combined information gained from the level of MRD at the post-induction and the pre-intervall period of therapy: 1. a low risk group which comprises 43% of all children with a relapse rate of 2%, 2. an intermediate risk group (43% of all children) with a relapse rate of 25-30%, and 3. a high risk group (14% of the children) with 75% relapses. Project aims: 1. To test whether the children in the next treatment protocol BFM ALL 99 will benefit from therapy adaptation tailored by the MRD results. Low risk patients will receive less intensive treatment and children in the intermediate risk group a new re-intensification treatment. These patients will be compared in a randomised trial with those receiving conventional therapy. High risk patients will be treated with other regimens than before (experimental treatment in extremely high risk children). 2. The MRD based intermediate risk group comprises the largest number of patients with a relapse (about 30% relapses in 43% percent of all children with ALL). Within this group there is currently no parameter which defines the patient at risk. We therefore will prospectively analyse the value of MRD levels from an additional bone marrow aspiration during intensification treatment. 3. We have previously experienced that the use of two independent antigen receptor gene rearrangements is necessary for the performance of a reliable MRD analysis. We therefore need to extend our current targets, i.e. immunoglobulin (Ig) k , T cell receptor (TCR) d , TCRg , and tall-deletion (for T-ALL) by including Ig heavy chain (IgH) gene rearrangement analysis. Thereby we will evaluate the two commonly used methods (FRI and FR3 PCR) with our recently developed touch-down FR3 PCR. The established collaboration with i) the laboratories of CR. Bartram (D) and A. Biondi (I) for the molecular methods, and ii) with the study center of the international BFM Study group M. Schrappe (D) for correlation of MRD results with clinical data will continue. Patients and Methods: We expect about 50 newly diagnosed children with ALL in Austria per year. The collaborating physicians have shown during the last MRD study that they are able to comply perfectly with the requested bone marrow aspirations (95% complete sampling). For the detection of MRD we will use the leukemia clone-specific junctional region of two antigen receptor gene rearrangements for the synthesis of oligonucleotides, which can be either used for dot blot hybridization of amplified DNA from follow-up samples or in a clone- specific PCR. A prerequisite for the selection of a target is high specificity and sensitivity for detection of at least one leukemic cell among 10.000 normal cells.

Background: Children with acute lymphoblastic leukemia are cured in about 75% with current treatment protocols. Earlier studies on minimal residual disease (MRD) - the presence of leukemic cells below the detection limit of morphological methods, i.e. 5% - have demonstrated that the amount of MRD during the first months of treatment correlates with prognosis. Aim of the study: The aim of this study was to test in an international collaboration (Germany, Italy and Austria) whether treatment stratification of all children included into the BFM ALL 99 Protocol can be based on MRD. After this feasibility phase, a randomized trial should show, whether patients with low risk for relapse according to MRD can be treated now in the 2000 Protocol with less intensive chemotherapy than before. Within the MRD-based high risk group children are randomized between different treatments in order to define the therapy, which results in higher cure rates. Further, day 15 and day 52 MRD should be evaluated regarding their impact on the definition of additional risk subgroups. Methods: In order to perform the analysis in a clinical study we adapted the MRD analysis to a real time quantification of leukemic cell load and included the incomplete and complete IGH rearrangements as targets, because such rearrangements are present in almost all B cell precursor leukemias. These methods were standardized and quality controls were performed among the collaborating laboratories. Results: We have demonstrated that in the collaborative study, which comprises now about 1000 patients, 80% of patients can be stratified for treatment according to MRD. Failure to get the MRD- based risk group assignment were inadequate samples, the presence of less than two targets, or the lack of the required sensitivity of the clone-specific PCR. There were about 35% in the MRD low-risk group, 55% in the MRD intermediate-risk group and 10% in the MRD high-risk group. Thus, the low- and high-risk groups were smaller compared to the previous study. Evaluation of relapses will be performed after an appropriate observation time (about 5 years after study start). We collected samples from about 200 patients from additional time points (day 15 and 52) for later analysis. Conclusion: The initial phase of the BFM ALL 1999 protocol demonstrates that it is feasible in a European study to monitor MRD in children with leukemias. Thus, we have created the prerequisite for the individualization of treatment, which may reduce treatment and thus toxicity and late effects in low-risk patients. On the other hand, the most intensive and also toxic treatment is given exclusively to the high- risk patients whose chance to be cured may raise with new treatment approaches. Further insight into the biology of leukemias was gained by the accompanying studies on MRD-monitored response in children with various subgroups of leukemias as well as on the number and types of leukemia clone-specific antigen receptor gene rearrangements.