Checkpoint Inhibition in CSF3R Mutated Chronic Neutrophilic Leukemia

Mutations in the colony-stimulating factor 3 receptor (CSF3R) have been identified in 90% of chronic neutrophilic leukemia (CNL) and lower frequency in atypical chronic myeloid leukemia (aCML). It was shown recently, that signaling of the different types of CSF3R mutations varies, but a central role of the JAK/STAT pathway activation by oncogenic CSF3R membrane proximal point mutations has been anticipated. However, the exact downstream signaling pathways mediated by CSF3R mutations are still not completely understood. Recent studies in JAK mutated myeloproliferative diseases uncovered the vital importance of JAK/STAT signaling in PD-L1 mediated immune escape, thereby suggesting a possible role and mechanistic link for PD-1/PD-L1 immune checkpoint regulation in CNL and aCML.

To investigate the dependence of the oncogenic CSF3R signaling cascade on an activated JAK/STAT axis, we performed Ruxolitinib treatment in MTT assays with BaF3 cells retrovirally infected with pMIG-R1 vectors harboring either CSF3R point or truncation mutations. Our data indicate, that both, CSF3R point and truncation mutations induced cell growth could be inhibited by Ruxolitinib treatment. To narrow down the involvement of the different JAK family kinases, we included CSF3R mutation transduced JAK1 deficient U4C, JAK2 deficient y2A and TYK2 deficient U1A cell lines in combination with different JAK inhibitors. Interestingly, here we were able to show that CSF3R point mutations require functional JAK1/TYK2-STAT3 signaling for oncogenic potential, whereas truncation mutations depend on all JAK kinases through STAT3/5 for activation.

To further study potential therapeutic strategies, we examined checkpoint inhibition in CSF3R mutated disease. In this regard we demonstrate a significant induction of PD-L1 expression in murine cell lines upon expression of CSF3R point mutations T615A or T618I. To investigate the impact of PD-L1 upregulation in vivo, we used a retroviral murine bone marrow (BM) transplantation model where we infected BM of Balb/c mice with a vector carrying CSF3R T618I point mutation and transplanted it into lethally irradiated recipients. Transplanted mice developed a CSF3R driven leukemia and died by leukocytosis and organ infiltration, whereas mice transplanted with BM harboring CSF3R^T618I and additional ADA mutation failed to induce leukemia. To test checkpoint inhibition as a potential therapeutic strategy in vivo, CSF3R^T618I transplanted mice were treated regularly with anti-PD-L1-antibody or isotype control. Interestingly, checkpoint inhibition led to reduced EGFP burden and prolonged survival of responder mice suggesting a treatment response of immunooncologic agents in these rare diseases.

In immunophenotypic analyses of primary patient material, we were able to show increased PD-L1 expression in primary diseased myeloid cells isolated from peripheral blood of CNL and aCML patients, thereby defining this structure as a possible new therapeutic target.

To summarize, our data expand the knowledge of oncogenic CSF3R signaling pathways and implicate an immunotherapeutic strategy by checkpoint inhibition in CSF3R driven CNL and atypical CML.