Acute lymphoblastic leukemia (ALL) in infants generally shows distinctive biologic features and has a poor prognosis. Cytogenetic studies indicate that many infant leukemias have chromosome 11q23 translocations. Because of these findings and the distinct clinical features of infant leukemia, we investigated 30 cases of infant ALL for molecular defects of 11q23. Fourteen cases had cytogenetic abnormalities of 11q23, and all of them showed 11q23 rearrangements at the molecular level. An additional seven cases also had 11q23 molecular rearrangements, including one with normal cytogenetic analysis. Molecular abnormalities of 11q23 were significantly correlated with adverse prognostic factors, including age under 6 months, hyperleukocytosis, CD10- phenotype, and early treatment failure. Molecular analysis identified a group of infants with germline 11q23 that had a very good treatment outcome with a projected event-free survival of 80% at median follow-up of 46 months compared to 15% in infants with rearranged 11q23 (P < .001). These findings suggest that a high proportion (70%) of infants with ALL have 11q23 rearrangements and that these rearrangements are not always detectable by cytogenetic analysis. The presence of germline 11q23 DNA may identify a subgroup of infant ALL patients with a good outcome using current therapy and a different etiology for their ALL.

The SCL gene, initially discovered at the site of a translocation breakpoint associated with the development of a stem cell leukemia, encodes a protein that contains the highly conserved basic helix-loop- helix (bHLH) motif found in a large array of eukaryotic transcription factors. Recently, we have described a nonrandom, site-specific SCL rearrangement in several T-cell acute lymphoblastic leukemia (ALL) cell lines that juxtaposes SCL with a distinct transcribed locus, SIL. The SIL/SCL rearrangement was found in leukemic blasts from 11 of 70 (16%) newly diagnosed T-cell ALL patients, a prevalence substantially higher than that of the t(11;14) translocation, which has previously been reported as the most frequent nonrandom chromosomal abnormality in T- cell ALL. We did not detect the SIL/SCL rearrangement in the leukemic blasts from 30 patients with B-cell precursor ALL, indicating that the rearrangement was specific for T-cell ALL. Analysis of RNA from these patients indicated that an SIL/SCL fusion mRNA was formed, joining SIL and SCL in a head-to-tail fashion. The fusion occurs in the 5′ untranslated region (UTR) of both genes, preserving the SCL coding region. The net result of this rearrangement is that SCL mRNA expression becomes regulated by the SIL promoter, leading to inappropriate SCL expression. The resultant inappropriate expression of this putative transcription factor may then contribute to leukemic transformation in T-cell ALL.

Monoclonal antibody L4F3 reacts with most acute myeloid leukemia (AML) cells and virtually all normal granulocyte/monocyte colony-forming cells (CFU-GM). Our objective was to determine whether lysis of AML cells with L4F3 and complement allowed expression of normal myeloid progenitors. The five glucose-6-phosphate dehydrogenase (G6PD) heterozygous patients with AML studied manifested only a single G6PD type in blast cells and in most or all granulocyte colony-forming cells, indicating that the leukemias developed clonally. The cells remaining after L4F3 treatment from two of the patients gave rise to granulocytic colonies that expressed the G6PD type not seen in the leukemic clone, indicating that they were derived from normal progenitors (CFU-GM). L4F3-treated cells from these two patients cultured over an irradiated adherent cell layer from normal long-term marrow cultures also gave rise to CFU-GM, which were shown by G6PD analysis to be predominantly nonleukemic. In the other three patients, the progenitor cells remaining after L4F3 treatment were derived mainly from the leukemic clone.