Malignant Transformation in Congenital Neutropenia

G-CSF is successfully used to treat severe congenital neutropenia (SCN), but a major complication is that SCN patients are at high risk to develop myelodysplastic syndrome (MDS) and acute myeloid leukemia (AML). Leukemic transformation correlates with the acquisition of nonsense mutations in the gene encoding the G-CSF receptor, now known as CSF3R. These mutations cause truncations of approximately 100 amino acids of the cytoplasmic domain of the receptor protein. Activation of CSF3R truncation mutants causes hyperproliferative responses to G-CSF in vivo and results in sustained STAT5 activation, an essential step in the clonal expansion of hematopoietic stem and progenitor cells (HSPCs) expressing these mutant receptors. Furthermore, activation of truncated CSF3Rs causes excessive production of reactive oxygen species (ROS). Conceivably, elevated ROS levels cause DNA damage, leading to an increased mutation rate in HSPCs. The time between the first detection of the CSF3R mutations and the diagnosis of MDS/AML varies between several months to more than 10 years. We performed whole exome sequencing on serial hematopoietic samples of SCN patients who progressed to MDS/AML. In a patient who developed AML 17 years after start of G-CSF treatment, 12 non-synonymous mutations were identified, two of which (in LLGL2 and ZC3H18) were acquired in a CSF3R-d715 mutant clone in the early SCN phase, i.e., 15 years before overt leukemia. This clone expanded in time, whereas clones harboring functionally similar but structurally distinct CSF3R mutations (d717, d725, d730) without the additional LLGL2 mutation became extinct. LLGL2 is a cell polarity protein and loss of LLGL activity has been associated with an increased self-renewal ability of HSPCs. Significantly, a new mutation (CSF3R-T595I) in the extracellular domain of CSF3R on the already affected CSF3R-d715 allele was present in the AML phase, resulting in autonomous proliferation of myeloid progenitors. These findings indicate that progression from SCN towards AML is a multistep process, in which clones with mutations arising early during the SCN phase (CSF3R-d715, LLGL2, ZC3H18) selectively expand and acquire additional mutations (e.g., SUZ12, ASXL1, RUNX1, CSF3R-T595I) at later stages during AML development. Whether CSF3R mutations are causative to leukemic progression of SCN and whether G-CSF treatment contributes to malignant transformation is unknown. Specifically, it remains to be resolved whether G-CSF merely drives the clonal expansion of the “sick” HSPCs or actively contributes to the leukemic process, e.g., through elevated ROS production after activation of the truncated CSF3R. The recent discovery of recurrent CSF3R mutations in chronic neutrophilic leukemia (CNL) sheds new light on the involvement of perturbed G-CSF signaling in the development of leukemia. The most prevalent mutation in CNL is identical to the auto-activating mutation (T595I) reported in SCN/AML, but named CSF3R-T618I (owing to the inclusion of the 23 amino acid signal peptide). Strikingly, about 30 percent of CNL patients with a T618I mutation also have a CSF3R truncation mutation present on the same allele. This suggests an intriguing parallel between the pathogenesis of SCN/AML and CNL, which may have clinical implications. In the context of CNL, it was shown that the tyrosine kinase inhibitor dasatinib can be applied to specifically inhibit clones expressing truncated CSF3Rs, whereas the auto-activating mutant CSF3R-T618I is sensitive to inhibition by ruxolitinib. We currently build models to explore how dasatinib and other inhibitors affect G-CSF signaling from wild type and truncated CSF3R. This will reveal how inhibitors can be employed in vivo to selectively eradicate SCN clones harboring truncated CSF3R to reduce the risk of malignant transformation.