Myeloproliferative neoplasms: from origins to outcomes

Abstract

Substantial progress has been made in our understanding of the pathogenetic basis of myeloproliferative neoplasms. The discovery of mutations in JAK2 over a decade ago heralded a new age for patient care as a consequence of improved diagnosis and the development of therapeutic JAK inhibitors. The more recent identification of mutations in calreticulin brought with it a sense of completeness, with most patients with myeloproliferative neoplasm now having a biological basis for their excessive myeloproliferation. We are also beginning to understand the processes that lead to acquisition of somatic mutations and the factors that influence subsequent clonal expansion and emergence of disease. Extended genomic profiling has established a multitude of additional acquired mutations, particularly prevalent in myelofibrosis, where their presence carries prognostic implications. A major goal is to integrate genetic, clinical, and laboratory features to identify patients who share disease biology and clinical outcome, such that therapies, both existing and novel, can be better targeted.

Figure 1.

Mutations in JAK2, CALR, and MPL drive excessive myeloproliferation via constitutively active signaling downstream of JAK2. JAK2 associates with the cytoplasmic portion of a variety of receptors, such as those for erythropoietin (EPOR), thrombopoietin (MPL), and granulocyte/macrophage colony-stimulating factor (G-CSFR). JAK2 is also activated in response to additional cytokines (eg, growth hormone and IL-5) (not shown). (A) Mutant JAK2, shown in red, is constitutively active and leads to variable levels of erythroid, megakaryocytic, and, to a lesser degree, granulocytic proliferation and differentiation. It is unclear whether mutant JAK2 dimerizes with mutant or wild-type JAK2 with respect to the individual receptors. (B) Mutations in CALR and MPL result in aberrant activation of signaling downstream of the MPL receptor. Mutant CALR complexes with MPL in the ER. Both mutations in CALR and MPL result in receptor dimerization and activation of JAK2. MAPK/ERK, mitogen-activated protein kinases/extracellular signal-regulated kinases; PI3/AKT, phosphoinositide 3-kinase/serine/threonine kinase Akt; STAT, signal transducer and activator of transcription.

Figure 2.

Clinical presentation in chronic phase and relationship to phenotypic driver mutation. PV and ET are modeled as a disease spectrum along a biological continuum where different genetic lesions skew the clinical phenotype from that of thrombocytosis to that of additional erythrocytosis (±leukocytosis). CALR mutations result in excessive MPL signaling, in a manner similar to that resulting from MPL mutations. JAK2 mutations signal downstream of multiple cell surface receptors, including MPL, and are thus associated with thrombocytosis but also erythrocytosis and leukocytosis. The exact nature of the phenotypic driver mutation, germline genetic background, and additional somatic mutations influence disease phenotype. *In the context of JAK2^V617F, several factors modulate the balance between erythrocytosis and thrombocytosis, including sex, mutation homozygosity, and patient-specific factors such as erythropoietin (EPO) levels, renal function, and iron status.

Figure 3.

From origins to outcomes. Different evolutionary paths to MPN and disease progression in 4 patients, each with a unique genetic background (green, gray, red, blue). In the first patient (green), a phenotypic driver mutation acquired in an HSC results in clonal expansion and the emergence of an MPN phenotype as a consequence of favorable cell-intrinsic and/or environmental factors. The MPN in this context has no additional oncogenic driver mutations, as is common for patients in chronic phase. Additional driver mutations, such as those that perturb polycomb repressor 2 function (EZH2, ASXL1 mutations), spliceosome components (SRSF2, SF3B1, U2AF1), or DNA damage repair (TP53), can lead to cells gaining a further clonal advantage and disease progression. In the second patient (gray), the cell-intrinsic and/or environmental context is not favorable, and a cell acquiring a phenotypic driver mutation does not have a clonal advantage relative to competing normal cells. In some circumstances, a phenotypic driver mutation may be insufficient to result in abnormal blood counts and an overt MPN but can instead result in a clonal expansion. Additional mutations or cell-extrinsic changes may be required to result in emergence of disease (patient in red). Finally, in some patients, phenotypic driver mutations may not be the first event. Clonal hematopoiesis as a result of mutations in, for example, TET2, DNMT3A, ASXL1 may be the required backdrop for a phenotypic driver mutation to result in an overt MPN (patient in blue).