Treatment with a MEK Inhibitor Improves Myeloproliferation, Anemia and Survival in a Mouse Model of CMML and JMML.

Abstract 966

Mutations that deregulate cellular signaling are a hallmark of myeloproliferative neoplasms (MPNs), and pharmacologic inhibitors of MPN-associated proteins have redefined therapy for some MPNs. However, this strategy cannot yet be applied to juvenile- and chronic myelomonocytic leukemias (JMML and CMML). These diseases are characterized by aberrant N-Ras, K-Ras, Cbl, and SHP-2 proteins that are not easily targeted by drugs. An attractive alternative approach is to inhibit downstream effector pathways, which include the Raf/MEK/ERK, phosphoinositide-3-OH kinase (PI3K)/Akt, and Ral-GDS/Ral-A cascades. However, it is not known which of these pathways are crucial for the aberrant growth and survival of JMML and CMML cells and might therefore provide the best targets for therapy. To address these questions, we developed an accurate mouse model of JMML and CMML by expressing a conditional “knock-in” Kras^LSL-G12D oncogene in bone marrow. We administered PD0325901, a potent and selective MEK inhibitor, to Mx1-Cre, Kras^G12D mutant mice to test the hypothesis that the Raf→MEK→ERK cascade is necessary for MPN initiated by Kras^G12D expression. Oral administration of PD0325901 5 mg/kg caused deep and durable MEK inhibition in primary bone marrow progenitors. Mx1-Cre, Kras^G12D mice with established MPN and wild-type (WT) littermates were randomly assigned to receive PD0325901 5 mg/kg/day or a control vehicle. Treated Mx1-Cre, Kras^G12D mice demonstrated rapid correction of leukocytosis and anemia, and reduction in splenomegaly. Treatment was also associated with dramatic improvement in the survival of Mx1-Cre, Kras^G12D mice (8.1 vs. 2.0 weeks after entry, p=0.003). Two of three Mx1-Cre, Kras^G12D mice that were treated for 12 weeks ultimately died with Kras^G12D T-lineage leukemia/lymphoma, but none succumbed with progressive MPN. Flow cytometry of bone marrow and peripheral populations showed that PD0325901 reversed the granulocyte/monocyte progenitor bias and ineffective erythropoiesis in Kras^G12D mice. However, PD0325901 did not eliminate the rearranged mutant Kras allele in myeloid progenitors, and these cells remained hypersensitive to GM-CSF in methylcellulose cultures. Therefore, PD0325901 did not eliminate Kras mutant cells, but rather modified their behavior in vivo so as to restore a normal output of the hematopoietic system. To further address the biologic effects of PD0325901 on growth of primary progenitor cells in vitro, we examined colony growth over a range of GM-CSF concentrations. Importantly, whereas in vitro exposure to PD0325901 did not selectively abrogate colony growth from bone marrow of naïve Mx1-Cre, Kras^G12D mice in the presence of saturating doses of GM-CSF, a low concentration of PD0325901 eliminated the growth of cytokine-independent progenitor colonies. Even more strikingly, this also restored a normal GM-CSF dose response curve in clonogenic progenitors, eliminating the hypersensitive growth pattern that is a hallmark of MPN. Finally, even at saturating doses of GM-CSF, a low concentration of PD0325901 was sufficient to normalize the numbers and types of cells within the colonies. Together, these data show that a low concentration of PD0325901 is sufficient to impart a normal program of proliferation and differentiation in Kras^G12D myeloid progenitors. These findings are highly consistent with the in vivo data. Collectively, our data suggest that aberrant MEK activation mediates most aspects of the MPN phenotype in the progenitor compartment and support the development of clinical trials to evaluate MEK inhibitors in patients with JMML and CMML.