Timing of genetic alterations in childhood ALL

The chromosomal translocation t(12;21) generating the ETV6/RUNX1 (E/R) fusion gene (also known as TEL/AML1) is the most frequent translocation in childhood acute lymphoblastic leukemia (ALL). Accumulating evidence suggests that the fusion gene formation is an early or even initiating event in leukemia development and occurs already in utero in the majority of cases with t(12;21) positive ALL. The fusion gene alone, however, is insufficient to cause overt leukemia. At least one, but, most likely, several other cooperating oncogenic lesions, as recently identified by Mullighan et al. (Nature 2007;446:758), need to occur in order to produce the clinical manifestation of leukemia. The timing and order in which these lesions emerge, however, are currently unknown. The aim of this study is to assess the timing of genetic lesions by tracking them back individually to archived neonatal blood spots, the earliest and frequently only blood sample available before the manifestation of leukemia. We will screen 50-60 ALL leukemia samples from children with t(12;21) positive ALL, from whom we have obtained the archived dried neonatal blood spots, for genetic alterations by genome-wide SNP arrays (6.0 SNP Affymetrix). Selected recurrent genetic lesions - preferably several lesions per individual case that belong to one of the major affected pathways, e.g. B cell differentiation and proliferation/cell cycle - will be cloned by a multiplex long-range PCR approach. Genomic DNA sequence information will be used to design highly specific and sensitive allele-specific PCRs for the ensuing screening of DNA from neonatal blood spots. If none of these genetic lesions can be detected in this material, the presence of the leukemia clone can be confirmed at birth by a clone- specific immunoglobulin (Ig) or T cell receptor (TCR) rearrangement or the patient-specific genomic E/R breakpoint. We expect these studies to provide a hierarchical order of secondary genetic lesions and thereby gaining new insight into the natural history of leukemia development. Furthermore we envision that the results facilitate a better characterization of persisting leukemic cells in the future, a prerequisite for novel treatment approaches.

Acute lymphoblastic leukemia (ALL) is the most common cancer in childhood and adolescence and the current cure rate approaches 80%. Yet despite this success, leukemia is still one of the leading causes of death in this age group. It is, therefore, essential to understand the development of leukemia in more detail to better treat, or perhaps even prevent this malignancy. Increasing evidence suggests that the vast majority of leukemia initiating events occur already in fetal life, but further complementary aberrations are required for the clinical manifestation of the disease. In this project we wished to test whether such secondary aberrations also develop in utero and, if so, whether they are generated in a hierarchical order. For this purpose we have chosen a specific subgroup of childhood leukemia that is characterized by the presence of the ETV6-RUNX1 fusion gene. The fusion gene is known to occur in utero in the vast majority of children with this leukemia subtype, as demonstrated by its detection in archived neonatal blood spots. Such blood spots are collected from all children after birth and used for metabolic disease screening; the remaining material is stored. As they often are the only blood source from before the outbreak of leukemia, we and others have used such dried blood spots to backtrack leukemia-associated markers to birth. Because the composition of secondary aberrations is distinct in each leukemia we first identified secondary aberrations by a genome-wide SNP array screening of the leukemias, then cloned selected breakpoints and used this information to design highly specific and sensitive assays that eventually allow detecting one aberrant among 100.000 normal cells. Altogether, we had 15 such aberrations from 10 leukemia children affecting critical genes in leukemia development and used them for the analysis of neonatal blood spot DNA. Despite the presence of the ETV6-RUNX1 fusion gene in all cases, none of the secondary aberrations were detected at birth. Our findings therefore support a scenario where secondary aberrations may occur only after birth. While this is conceivable, our data do not exclude the possibility of an in utero origin, because the numbers of cells that are contained on one neonatal blood spot are very limited and only 10 to 100 of them harbor the leukemia-specific fusion gene. Moreover, recent data suggest that the individual leukemia is not a clonal disease, as previously assumed, indicating that it contains several cell populations with various combinations of aberrations, which are generated during the leukemogenic process that involves several rounds of mutations and selections. Thus, the selected aberrations that prevail in leukemia at diagnosis might only be present in a small proportion of ETV6-RUNX1-positive cells at birth. These considerations to define the dignity of such cells at birth are not only of biological interest but also key to the preservation of cord blood cells, because it is important to know whether they can be safely used for reinfusion without an increased risk for cancer development.