Enhanced Molecular Response in Myeloproliferative Neoplasms with Complete JAK2^V617F Inhibition

Myeloproliferative neoplasms (MPN) represent a spectrum of clonal hematopoietic stem cell disorders characterized by pathologic overproduction of erythrocytes and thrombocytes in polycythemia vera (PV) and essential thrombocythemia (ET), respectively, and accumulation of extracellular matrix in the bone marrow (BM) in myelofibrosis (MF). The oncogenic point mutation JAK2^V617F occurs in more than half of all patients with MPN (1). The V617F mutation disrupts the inhibitory activity of the JAK2 pseudokinase domain, leading to constitutive JAK2 kinase activity, increased production of proinflammatory cytokines, and hyperactivation of multiple downstream progrowth and prosurvival pathways (2). The acquisition of cooperating mutations and the order in which these mutations are acquired influences the clinical characteristics of the disease, how patients respond to targeted therapies, and rates of secondary acute myeloid leukemia transformation (3). Because of the initial description of this mutation in patients with MPN, there has been tremendous activity toward the development of inhibitors for clinical application. Unlike most hematopoietic neoplasms, JAK2^V617F is sometimes the only pathogenic mutation driving the disease in patients with MPN, suggesting effective inhibitors could be curative for some patients. Four JAK2 inhibitors have received FDA approval for treating patients with MPN as of 2023—ruxolitinib, fedratinib, pacritinib, and momelatinib, which all exhibit differential clinical responses because of their effects in targeting other kinases beyond JAK2 (4). All four of these agents are classified as type-I JAK2 inhibitors that primarily bind to the ATP-binding site of JAK2 in its active, phosphorylated state. In contrast, type-II inhibitors engage not only the ATP-binding site but also the surrounding regions when JAK2 is in an inactive conformation. Despite the clinical success of these drugs in alleviating constitutional symptoms, reducing splenomegaly, and improving the quality of life for patients, these inhibitors do not eliminate the malignant clone and have a modest impact on BM fibrosis and overall survival (5). The suboptimal efficacy of current JAK2 inhibitors is hypothesized to arise from a combination of factors: (i) inability to fully repress oncogenic activity of the JAK2^V617F variant; (ii) acquisition of additional pathogenic mutations reduces the dependency on JAK2^V617F for disease maintenance; and (iii) nonselectivity of these pan-JAK2 inhibitors for the JAK2^V617F variant results in unintended off-target effects like anemia, leading to early treatment discontinuation. This has brought into question whether the JAK2^V617F variant itself is the most appropriate therapeutic target in chronic patients with MPN.

Several transgenic and conditional knock-in mice have been developed to model JAK2^V617F–driven MPN that recapitulate the molecular and clinical characteristics of patients with MPN to varying degrees (6). These models have established that the JAK2^V617F mutation is acquired at the level of hematopoietic stem cells (HSC), and this mutation alone can induce a PV- or ET-like disease (6). Although these current models have helped characterize the pathobiology of MPNs and identified novel therapeutic targets (7), they do not allow an accurate assessment of the dependence of MPN HSCs on JAK2^V617F, either alone or in combination with cooperating mutations that arise during clonal evolution and leukemic transformation.

To overcome the limitations of current mouse systems, Dunbar, Bowman, and colleagues have developed an elegant dual-recombinase animal model of MPN that enables temporal control over expression and repression of the JAK2^V617F mutation in physiologica context (8). This is achieved through a dual recombinase system in which Dre-rox recombinase triggers induction of the JAK2^V617F in the endogenous locus, leading to a highly penetrant MPN in primary and secondary transplants (8). Recipient animals transplanted with hematopoietic stem and progenitor cells (HSPC) expressing Dre recombinase developed PV-like disease, including leukocytosis, erythrocytosis, hepatosplenomegaly, and megakaryocytic hyperplasia, with a significant proportion progressing to reticulin fibrosis in long-term or serial transplantation. These phenotypes are consistent with the previously established mouse models of MPNs.

However, the elegance of this model is that once the disease is established, the mutant JAK2^V617F allele can be inactivated by Cre-lox recombination, here induced with tamoxifen in a tissue-specific manner. In noncompetitive transplants, tamoxifen treatment resulted in near complete inactivation of JAK2^V617F, subsequently leading to the elimination of the MPN disease. This was evidenced by the normalization of white blood cell, hematocrit, and platelet counts within four weeks that was sustained long-term and induced resolution of disease pathologies. Notably, inactivation of JAK2^V617F induced acute apoptosis in the pathogenic HSPCs, reinforcing the notion that the disease-initiating cells are dependent on JAK2^V617F for their survival, which validates this as the correct target for therapeutic strategies. A small number of animals did experience disease recurrence, but this was attributed to the incomplete excision of the JAK2^V617F allele, which further underscores the absolute reliance of the disease phenotype on this mutation. Retransplantation of JAK2^V617F-deleted BM failed to reinitiate disease, consistent with the depletion of disease-propagating MPN stem cells upon JAK2^V617F inactivation. Transcriptional profiling of HSPCs following acute JAK2^V617F depletion revealed reduced expression of prosurvival pathways, including MAPK and MTORC1, as well as proinflammatory signaling through STAT3/5, IFNγ, TGFβ, and TNFα, which are the molecular hallmarks of MPNs.

Dunbar, Bowman, and colleagues next compared the phenotypic effects of JAK2^V617F deletion to type-I inhibition by ruxolitinib or type-II inhibition by CHZ868, the latter of which demonstrated improved potency and slight selectivity for JAK2^V617F compared with ruxolitinib (9). Treatment with ruxolitinib (a pan-JAK1/2 dual inhibitor) did not result in the same extent of disease improvement as genetic deletion of JAK2^V617F in this new model of MPN. In contrast, treament with CHZ868 produced symptom benefits more on par with JAK2^V617F deletion, including a reduction in mutant allele burden in peripheral blood and megakaryocytic progenitors in the BM (Fig. 1). This finding is incredibly important, as the inability to achieve complete disease eradication or more substantial molecular and hematologic responses in patients with MPN treated with JAK2 inhibitors might stem from the insufficient and/or nonselective inhibition of the pathogenic JAK2^V617F rather than reduced dependence of MPN cells on JAK2^V617F. These data underscore the critical need for more potent and selective targeting agents, such as new type-II JAK2 inhibitors or JH2-JAK2^V617F inhibitors currently in preclinical development.

Genetic alterations in the TET2 gene occur in 7%–13% of patients with MPN. Functional studies have demonstrated that JAK2^V617F and TET2 mutations can act synergistically to improve the fitness of pathogenic HSCs and increase the risk of disease progression. Furthermore, the order of acquisition of these mutations in HSCs has been shown to influence the response to resulting MPN to therapy (10), suggesting each mutation may prime the premalignant HSC using particular molecular mechanisms prior to acquisition of the second pathogenic mutation. This model from Dunbar, Bowman, and colleagues provides an ideal experimental system to mimic the sequential acquisition of mutations and then assess the dependency of the resulting disease-initating cells on JAK2^V617F for MPN maintenance. Consistent with previous studies, the induction of JAK2^V617F in a TET2-deficient background led to more pronounced MPN disease pathologies compared with JAK2^V617F alone in a transplant setting. But remarkably, the deletion of JAK2^V617F in the double-mutant cells led to almost complete normalization of all MPN features. This included the loss of pathogenic HSC self-renewal capacity and reversal in BM fibrosis similar to the levels observed with single mutant JAK2^V617F deletion. The data collectively suggests that TET2 mutations acquired before JAK2^V617F do not alter the reliance of the disease-initaiting cells on JAK2^V617F signaling and that targeting JAK2^V617F is a viable clinical approach even in the context of other additional cooperating mutations.

In summary, this novel model definitively demonstrates there is an absolute requirement for JAK2^V617F in MPN disease-initiating cells for survival, and targeting with more potent and selective JAK2^V617F inhibitors could represent a potential curative strategy for the treatment of patients with MPN, even in the presence of additional cooperating mutations.