Sequential Acquisition of Somatic Mutations In Progenitors with Differential Proliferative Potential In Human Acute Myeloid Leukemia

Evidence suggests that leukemic transformation is a multistep process in acute myeloid leukemia (AML). However, the cell of origin for human leukemogenesis remains ill defined. Experimental systems with murine hematopoetic cells indicate that not only immature multipotent hematopoietic stem cells but also more mature, committed myeloid precursors can be a target of murine AML transformation. These findings suggest that requisite mutations for leukemogenesis may be acquired sequentially in myeloid progenitors with differential proliferative potential. To investigate this possibility in human AML, we studied primary specimens from pediatric patients who had previously been identified as having 2 distinct, disease-relevant somatic mutations.

Diagnostic bone marrow or peripheral blood specimens from pediatric AML patients with two distinct genomic alterations were obtained from Children's Oncology Group (COG) AML Reference Laboratory. Early myeloid CD34+/CD33- progenitors and more differentiated CD34+/CD33+ progenitors were isolated using fluorescence-activated cell sorting (FACS). Twenty-eight of 43 patient specimens had FACS-isolated CD34+/CD33- progenitor cells and molecular data available for both somatic alterations, and were included in this study. The genomic alterations included those with core-binding factor (CBF) translocations and FLT3/ITD (n=5), CBF translocations and KIT abnormalities (n=19), or nucleophosmin (NPM1) mutations and FLT3/ITD (n=4). Aliquots of sorted subpopulations were subjected to direct colony forming cell (CFC) assays. Freshly sorted cells and resulting colony forming units-granulocyte and/or monocyte (CFU-GM) were evaluated for the presence of the patient-specific genomic alterations.

In 20/28 specimens (71.4%), both mutations were detected in CD34+/CD33- cells. By comparison, in 6 (21.4%) patient specimens, the receptor tyrosine kinase (FLT3/ITD or KIT) alteration was not detected in CD34+/CD33- cells, whereas in 2 (7.1%) specimens, CD34+/CD33- cells did not harbor either of the two patient-specific alterations. In contrast to these less mature CD34+/CD33- progenitor cells, analysis of more differentiated CD34+/CD33+ progenitors revealed both genomic alterations in 26/28 samples. For the remaining 2 samples, only the CBF mutation was detected. FACS-isolated CD34+/CD33- cells from 22 specimens were subjected to CFC assay. There was sufficient colony growth and evaluable molecular data for the paired mutations for 7/22 samples. Within these specimens, the majority (43 of 60) of the evaluable CFU-GM demonstrated the CBF or NPM1 abnormality but lacked the RTK mutation; in the remaining colonies, both somatic alterations were detected (10 of 60), or neither mutation was detected (7 of 60). This finding suggests that the majority of CFU-GM analyzed were derived from an abnormal cell that had not yet acquired the requisite RTK mutation necessary for frank leukemic transformation.

These findings suggest the potential for sequential acquisition of RTK mutations in some cases of pediatric CBF or NPM1+ AML. Specifically, the presumptive class I mutation confering a proliferative advantage (e.g. the receptor tyrosine kinase mutation) appears to arise, in some instances, in a more committed myeloid progenitor with limited proliferative potential.

No relevant conflicts of interest to declare.