Key points

Introduction

Myeloproliferative neoplasms (MPNs) are chronic leukemias occurring at an annual incidence of 0.5 to 1.0 per 100,000. They are hematopoietic stem cell disorders leading to excessive proliferation of mature myeloid cells. The 3 main MPN clinical phenotypes include essential thrombocythemia (ET) with marked thrombocytosis, polycythemia vera (PV) with erythrocytosis often along with neutrophilia and thrombocytosis, and myelofibrosis (MF) with expansion of megakaryocytes, progressive bone marrow fibrosis, and extramedullary hematopoiesis. Splenomegaly and elevated serum cytokine levels are typical and contribute to symptom burden. Myelofibrosis is the rarest (0.47/100,000 annually ) and most severe form of MPN with life expectancies limited to a few months in the presence of high-risk features. MF transforms to secondary acute myeloid leukemia (AML) in 15% to 20% of patients (0.09/100,000 annually) with dismal prognosis or alternatively leads to hematopoietic failure. The sole curative therapy for MPN to date is hematopoietic stem cell transplantation. As MPNs mainly affect the elderly, only a limited subset of patients is eligible showing reduced success rates. Interferon-α has shown disease-modifying activity with molecular responses, but application is limited by poor tolerability. Other therapies focus merely on symptom control and prevention of thromboembolic and hemorrhagic events complicating the disease, without altering the natural course of MPN.

The breakthrough discovery of the V617F gain-of-function mutation in the tyrosine kinase Janus kinase 2 ( JAK2 ) in 2005 for the first time denominated a molecular therapeutic target. JAK2 V617F is present in 95% of PV and 50% to 60% of MF and ET. JAK2 is an intracellular non-receptor tyrosine kinase essential for hematopoiesis. It represents the exclusive mediator of cytokine signaling from the thrombopoietin receptor MPL, the erythropoietin and granulocyte-macrophage colony-stimulating factor receptors. Hematopoietic cytokines binding to their cognate receptors induce JAK2 dimerization and phosphorylation. Activated JAK2 initiates signaling through several intracellular signaling pathways, including the signal transducers and activators of transcription (STAT)3 and STAT5 transcription factors, the phosphoinositide-3 kinase (PI3K)/Akt pathway, and the mitogen-activated protein kinase (MAPK) pathway, which promote cell proliferation, differentiation, and survival via multiple effectors. Thus, the JAK2 V617F mutation constitutively activates JAK2 signaling leading to dysregulated myeloid cell proliferation in most patients with MPNs. Further molecular characterization identified JAK2 exon 12 mutations in most cases of JAK2 V617F-negative PV and mutations in the thrombopoietin receptor MPL, such as MPL W515 L and MPL W515 K in JAK2 V617F-negative patients with ET/MF, accounting for 5% to 10% of ET and MF. As JAK2 V617F, these mutations induce hyperactive JAK2 signaling and uncontrolled growth of myeloid lineages. Recently, acquired mutations in the chaperone protein calreticulin (CALR ) were identified in most patients with ET/MF who were JAK2 -unmutated and MPL -unmutated, accounting for 25% to 35% of ET and MF. There is mechanistic evidence that CALR mutations converge on activation of MPL through facilitating binding of mutant CALR to MPL. Thus, mutations in JAK2, CALR, and MPL , as well as rare mutations in negative regulators of JAK2, such as LNK and CBL , all induce activated JAK2 signaling. They are mutually exclusive, which highlights the JAK2 pathway activation as a shared mechanism of transformation in MPNs. Transcriptional profiling of MPN granulocytes substantiated this notion, showing gene expression profiles consistent with activated JAK2 signaling in all patients with MPNs independent of mutational status and clinical phenotype. Additional genetic lesions, such as mutations in epigenetic modifiers TET2 , ASXL1 , EZH2, or IDH1/2, or in tumor suppressors, such as TP53 , have also been identified in MPNs and may shape the heterogeneous phenotypes of MPNs and impact the course, whereas JAK-STAT pathway activation represents the central mechanism in the pathogenesis of all MPNs.

The critical role of hyperactive JAK2 signaling in MPN has led to the development of JAK2 inhibitors. Conventional JAK2 inhibitors act as ATP mimetics and stabilize JAK2 in the active conformation characterized by paradoxic hyperphosphorylation of the JAK2 activation loop (type I JAK inhibition). They effectively reduce splenomegaly and constitutional symptoms in patients with MPN, which means substantial alleviation of symptom burden. A survival advantage has been observed compared with placebo or best available therapy. Type I JAK inhibitors are not mutant-selective, meaning that they inhibit both mutant JAK2 as well as activated wild-type JAK2 downstream of mutated MPL or CALR , providing effective treatment in JAK2 mutant and JAK2 wild-type patients with MPN. Ruxolitinib, a JAK1/JAK2 inhibitor has been approved in the United States and Europe for treatment of intermediate-risk and high-risk MF and for PV resistant or intolerant to hydroxyurea. Similar compounds are in clinical development. Momelotinib, a JAK1/JAK2 inhibitor, is currently being compared with ruxolitinib in a phase III study in MF ( NCT01969838 ) and has led to transfusion independence in a significant proportion of patients in a previous phase I/II study. BMS911543 was pursued as a more JAK2-selective inhibitor and completed a phase I/II study ( NCT01236352 ), while additional compounds with JAK2/FLT3 inhibitory profile (fedratinib, pacritinib) did not complete phase III studies due to occurrence of adverse events. Overall, type I JAK2 inhibitors represent a major step forward for the treatment of MPNs, but they have not met the high expectations. A convincing body of evidence demonstrates that they are unable to substantially repress the mutant clone. This inability to induce significant reductions of mutant allele burden suggests a limited curative potential. Moreover, type I JAK inhibitors induce resistance on prolonged exposure, and we also observed cross-resistance among several type I JAK inhibitors in clinical development. Given the limited curative potential of type I JAK2 inhibitors, extended duration of treatment is critical to achieve long-term disease control with these agents. Therefore, a detailed understanding of resistance mechanisms is very relevant to inform the development of superior treatment strategies for MPN. Innovative therapeutic approaches to overcome resistance to type I JAK inhibitors are needed.