Wiley MPD 2009

Advances in Understanding and Management ofMyeloproliferative Neoplasms

Alessandro M. Vannucchi, MD1; Paola Guglielmelli, MD2; Ayalew Tefferi, MD3

AbstractAccording to the 2008 World Health Organization classication system for hematologic malignancies, themyeloproliferative neoplasms (MPN) include chronic myelogenous leukemia, polycythemia vera, essentialthrombocythemia, primary myelobrosis, mastocytosis, chronic eosinophilic leukemia-not otherwise specied,chronic neutrophilic leukemia, and MPN, unclassiable. All of these clinicopathologic entities are characterized bystem cell-derived clonal myeloproliferation, and their phenotypic diversity is ascribed to the occurrence of distinctoncogenic events. In the last 4 years, new JAK2 and MPL mutations have been added to previously described ABLand KIT mutations as molecular markers of disease in MPN. These discoveries have markedly simplied theapproach to clinical diagnosis and have also provided molecular targets for the development of small-moleculedrugs. In the current article, the authors provide a clinically oriented overview of MPNs in terms of their molecularpathogenesis, classication, diagnosis, and management. CA Cancer J Clin 2009;59:171-191. 2009 AmericanCancer Society, Inc.

To earn free CME credit or nursing contact hours for successfully completing the online quiz based on this article, goto http://CME.AmCancerSoc.org.

IntroductionIn 1951, William Dameshek1 introduced the term myeloproliferative disorders (MPD) to encompass polycy-themia vera (PV), essential thrombocythemia (ET), primary myelobrosis (PMF),2 chronic myelogenous leuke-mia (CML), and Di Guglielmos syndrome (erythroleukemia). His proposal was based on similarities in theclinical phenotype of these disorders and on the hypothesis that a generalized proliferation of bone marrow cells,due to some unknown stimuli, was the underlying cause. The association of the Philadelphia (Ph1)-chromosomewith CML in 1960,3 and the subsequent recognition of erythroleukemia as a variant of acute myeloid leukemia(AML), distinguished the other three disorders as classic Ph1-negative MPD.4

The rst systematic attempt to classify MPD and MPD-like clinicopathologic entities was undertaken by theWorld Health Organization (WHO) committee for the classication of hematologic malignancies.5 Accordingto the 2001 WHO classication system, CML, PV, ET, and PMF were included under the category of chronicmyeloproliferative diseases (CMPD). The CMPD category also included other nonclassic MPD-like disorderssuch as chronic neutrophilic leukemia (CNL), chronic eosinophilic leukemia/hypereosinophilic syndrome (CEL/HES), and unclassied CMPD. The identication of BCR-ABL as a CML-specic genetic event, in thecontext of CMPD, has facilitated accurate molecular diagnosis and effective targeted therapy. The lack of

1Associate Professor of Hematology, Department of Hematology, University of Florence, Florence, Italy. 2Research Fellow at the Department of Hematology,University of Florence, Florence, Italy. 3Professor of Medicine and Hematology, Mayo Clinic College of Medicine, Rochester, NY.

Corresponding authors: Alessandro M. Vannucchi, MD, Hematology Unit, Dip. Area Critica, University of Florence, Viale Morgagni 85, 50134 Florence, Italy;[email protected] and Ayalew Tefferi, MD, Division of Hematology, Mayo Clinic College of Medicine, 200 First Street SW, Rochester, MN 55905;[email protected]

DISCLOSURES: This study was supported by Associazione Italiana per la Ricerca sul Cancro, Milano; Istituto Toscano Tumori; MIUR (COFIN 2006067001_003). Theauthors report no conicts of interest.

2009 American Cancer Society, Inc. doi:10.3322/caac.20009.

Available online at http://cajournal.org and http://cacancerjournal.org

CA CANCER J CLIN 2009;59:171-191

171VOLUME 59 NUMBER 3 MAY/JUNE 2009

knowledge, until recently, onspecic genetic defects in theother BCR-ABL-negative classicCMPDs necessitated that diag-nosis rest on a combination ofbone marrow histology and afew clinical and laboratory nd-ings to distinguish clonal fromreactive myeloproliferation andone CMPD from another.6

The last 4 years have wit-nessed fundamental advances inunderstanding the molecularpathogenesis of classic BCR-ABL-negative CMPD, cappedby the discovery of specic mo-lecular abnormalities associatedwith PV, ET, and PMF.7 As aresult, WHO diagnostic criteriahave been revised,8 and the termCMPD has been changed tomyeloproliferative neoplasms(MPN).8 It is hoped that newlydiscovered mutations will also fa-cilitate development of targeted therapy. At the sametime, large clinical studies continue to provide practi-cally useful clinical information.

The current review has two main objectives. The rstis to provide a general overview ofMPN including theirmolecular pathogenesis and updated WHO classica-tion. The second objective is to describe in more detailthe criteria for diagnosis, risk stratication, and man-agement of patients with the classic BCR-ABL-nega-tive MPN including PV, ET, and PMF.

Molecular Basis of MyeloproliferativeNeoplasmsApart from the BCR/ABL rearrangement in CML,originated by a reciprocal translocation betweenchromosomes 9 and 22, t(9;22)(q34; q11),9 or thechimeric FIP1L1-PDGFRA mRNA in some formsof eosinophilia,10 and kit mutations in cases withsystemic mastocytosis,11 information concerning mo-lecular abnormalities of MPN has been scanty until2005, when a Janus kinase 2 mutation (JAK2V617F)was discovered in the majority of patients with PVand in 50% or fewer of those with ET or PMF.12-15

In the following 2 years, additional mutations inJAK216 and MPL17,18 were reported (Table 1). Thesedifferent mutant alleles all result in a gain of functiondue to the constitutive activation of tyrosine kinase-dependent cellular signaling pathways, particularly ofthe JAK-STAT pathway.19,20 Overall, this wouldsuggest that mutated kinases represent a commonpathogenetic mechanism in these disorders and that,as exemplied by the efcacy of the tyrosine kinaseinhibitor imatinib in CML, they could representvalid targets for therapy.21,22

Members of the Janus kinase family (JAK1, JAK2,JAK3, and tyrosine kinase 2-Tyk2) are named afterthe Roman god with two faces, meaning ending andbeginning, because they contain two symmetrical ki-nase-like domains: the C-terminal JAK homology 1(JH1) domain possesses tyrosine kinase function,whereas the immediately adjacent JH2 domain isenzymatically inactive, but it is credited with nega-tively regulating the activity of JH1.23,24 Ordinarily,JAKs are associated in an inactive state to the cyto-plasmic tail of type 1 or type 2 cytokine receptors (eg,erythropoietin receptor, EpoR; thrombopoietin re-ceptor, MPL; granulocyte colony-stimulating factorreceptor, G-CSFR; and interferon-gamma receptor,

TABLE 1. Recurrent Molecular Abnormalities Associated with MyeloproliferativeNeoplasms

GENETIC ABNORMALITY DISEASE FREQUENCY

BCR-ABL Chronic myelogenous leukemia 99%

JAK2V617F Polycythemia vera 95%

Essential thrombocythemia 60%

Primary myelobrosis 60%

MPN, unclassiable 20%

Refractory anemia with sideroblasts andthrombocytosis (RARS-T)

50%

JAK2 exon 12 Polycythemia vera 2%

MPLW515L/K* Primary myelobrosis 8%

Essential thrombocythemia 8%

Involving PDGFRA Myeloid neoplasms with eosinophilia Unknown

Mast cell disease Unknown

Involving PDGFRB Myeloid neoplasms with eosinophilia Unknown

Involving FGRF1 Myeloid neoplasms with eosinophilia Unknown

Involving KIT (D816V as the most frequent) Mast cell disease Unknown

MPN indicates myeloproliferative neoplasm.*Other infrequent mutations, such as W515A or S505N, have been reported.Calculated on JAK2V617F-negative patients.

Myeloproliferative Neoplasms

172 CA: A Cancer Journal for Clinicians

to name a few). After the engagement of the receptorby corresponding ligand, JAK undergoes a confor-mational change and becomes activated via phospho-rylation of key tyrosine residues. In turn, phospho-rylated JAKs mediate phosphorylation of tyrosineresidues of the cytoplasmic domain of the receptorsand create a docking site for the recruitment ofseveral proteins, ultimately leading to activation ofthe signal transducer and activator of transcription(STAT), the mitogen-activated protein (MAP) ki-nase, and the phosphatidylinositol 3-kinase-AKT(PI3K-AKT) pathways25 (Fig. 1A). ActivatedSTATs dimerize and translocate to the nucleuswhere they regulate transcription after binding tospecic consensus sequences in the promoter regionsof several target genes (Fig. 1A). The entire processis tightly controlled at multiple levels by proteintyrosine phosphatases, suppressors of cytokine sig-naling (SOCS), and protein inhibitors of activatedSTAT.26-29 JAK2, and possibly other JAKs, is alsoinvolved in the expression of cognate receptorsEPOR and MPL at the cell surface by acting as achaperon and protein stabilizer.30,31

The JAK2V617F mutation is a somatically ac-quired G to T nucleotide shift at position 1849 inexon 14 that results in a valine to phenylalaninesubstitution at codon 617; the mutation is located inthe JH2 pseudo-kinase domain and is believed toresult in the loss of auto-inhibitory control of JAK2(Fig. 1B). As a consequence, mutated JAK2 is in aconstitutively phosphorylated state, independentfrom the binding of ligand to its receptor; in fact,when the mutation is introduced into cytokine-de-pendent cell lines it results in a cytokine-independentgrowth of the cells and their hypersensitivity to cy-tokines,13,14 mimicking the in vitro growth pattern ofhematopoietic progenitors from MPN patients. Inparticular, the gain of function of mutated JAK2provides a mechanistic explanation for the phenom-enon of endogenous erythroid colony formation(EEC),33,34 ie, the capacity of erythroid progenitorsto spontaneously produce hemoglobinized coloniesin vitro in the absence of added erythropoietin, ahallmark of PV and other classic MPNs. Further-more, transplantation of JAK2V617F mutated cellsinduced a PV-like phenotype in recipient mice,13,35-38

accompanied by leukocytosis of a different extent andeventually followed by changes suggestive of myelo-brotic transformation.35-38 More recently, by ma-

nipulating expression levels of the V617F allele, micewith an ET-like phenotype were also generated inthe presence of low levels of mutated JAK2.39 Over-

FIGURE 1. (A) In normal hematopoietic cells, signaling is initiated when cyto-kines bind to and activate their cell surface type-1 receptors, which havemoleculesof JAK2 associated to the cytoplasmic domains. After ligand engagement (thepathway activated by EPO bound to the EPOR is herein schematized) the receptor-associated JAKsbecomeactivated throughauto-phosphorylation andon turnphos-phorylate tyrosine residues in the receptor cytoplasmic tail. The receptor phospho-tyrosines serve as docking sites for the recruitment of inactive cytoplasmic STATmonomers through interaction with their SH2 domain. JAK-mediated phosphoryla-tion of tyrosine residues on the receptor-bound STAT monomer induces STATsdimerization. The activated dimers translocate to the nucleus, where they bind tospecic DNA-responsive elements in the promoters of target genes and therebyinduce unique gene expression program(s). Activation of JAK2 pathway also resultsin the recruitment and activation of MAPk signaling proteins and AKT/mTOR/FOXOpathway that transmit signals for survival, proliferation, and differentiation oferythroblastic progenitors; JAK2-independent activation of these pathways mightalso occur. Negative feedback mechanisms are normally mediated, among otherregulators, by SOCS proteins. These complex signals are autonomously activated,in the absence of binding of the cytokine to its receptor, when JAK2 is mutated(JAK2V617For activatingmutations in exon12) or the receptor itself ismutated (asis the case of W515L/K mutation of MPL receptor). (B) Schematic representationof the most common genetic abnormalities associated with MPN. For details,please refer to text.STAT indicates signal transducer and activator of transcription; AKT, protein kinaseB, PKB; FOXO, forkhead transcription factors; PI3K, phosphatidylinositol-3-kinase;MAPK, mitogen-activated protein kinase; mTOR, mammalian target of rapamycin;SOCS, suppressor of cytokine signaling; JAK2, Janus kinase 2 gene;MPL, thrombo-poietin receptor gene; FIP1L1-PDGFRA, fusion gene of Fip1-like 1 with platelet-derived growth factor receptor alpha; kit: stem cell factor (SCF) receptor gene;FERM: 4-point-1, Erzin, Radixin, Moesin JAK2 amino-terminal domain; JH1, JAKhomology 1 (active tyrosine kinase) domain; JH2, JAK homology 2 (catalyticallyinactive pseudokinase) domain; SH2, SRC homology 2 domain; HRD1, HRD2, He-matopoietin/cytokine receptor domain1 (negative regulatory domain) or domain2(ligand binding region); SP, signal peptide; TM, trans-membrane domain; JM, jux-tamembrane domain; Ig, Immunoglobuline-like repeat; K1, Kinase domain 1; KI, 76amino acids kinase insert domain; K2, kinase domain 2.



all, these models indicated that the JAK2V617F mu-tation is sufcient to induce a MPN-like phenotypein mice and suggested that the level of mutated allelemay inuence disease phenotype.40

Mutational frequency of JAK2V617F is estimatedto be more than 95% in PV, 60% in ET or PMF,40% to 50% in refractory anemia with ringed sidero-blasts and thrombocytosis (RARS-T),41 whereas it isvery rare in AML or MDS.42-45 In most patients withPV or PMF, as opposed to a minority of those withET, the mutation is harbored in a homozygous state,which is accomplished by mitotic recombination.12-15

In general, the highest V617F allele burden, that is thelevel of mutated allele relative to normal allele in a cellsuspension such as granulocytes, is found in patientswith PV followed by PMF and ET46,47; however, suchvariability in the allele burden does not represent asufcient criterion for distinguishing among differentclinical entities, nor does it satisfactorily help to explainthe apparent paradox of one mutant allele-differentclinical phenotypes. In fact, how a single V617F mu-tation can be the basis of different clinical disorders, asin the classic MPN, is still unclear. Interestingly, singlenucleotide polymorphisms (SNPs) in JAK2 have beenassociated preferentially with the diagnosis of PV,48

supporting the contribution of inherited host geneticcharacteristics to MPN phenotypic variability. Regard-less, there is evidence to suggest that JAK2V617F maynot be the initial clonogenic event in MPN and that apre-JAK2 mutated cell may exist.49,50 In support ofthis is also a nding that leukemic blasts in patients whoevolve to AML from a pre-existing JAK2V617F-posi-tive MPN are often negative for the JAK2V617F mu-tation.51,52 Conversely, JAK2V617F, or other JAK2mu-tations, are likely a necessary component of the PVphenotype because they are detected in virtually allpatients with the disease53 and are sufcient to repro-duce the phenotype in mice. In summary, JAK2V617Fmutation is integral to the classic MPN, but its exacthierarchical position in pathogenesis and its role inphenotypic variability remain to be claried. After all,one could conclude that PV, ET, and PMF are separatediseases or different presentations or different phases ofa unique disease. It has been suggested that the pheno-type of patients with JAK2V617F-positive ET resem-bles forme fruste of PV.54

In patients with a clinical picture suggestive of PVand who were found to be negative for theJAK2V617F mutation, several genetic abnormalities

(ie, mutations, deletions, insertions) have been de-tected in a short region of JAK2 exon 12 (Fig.1B).16,55 These mutations, which probably accountfor less than 2% of patients with PV,55 affect auton-omous cell proliferation and differentiation in a fash-ion similar to that of the V617F allele.16

Another recurrent molecular abnormality of MPNis represented by somatic mutations at codon 515 ofMPL,17,18 which, as is the case with JAK2V617F,involve early myeloid and lymphoid progenitors.56-58

MPL (named after myeloproliferative leukemia virusoncogene homolog) is the receptor for the cytokinethrombopoietin (Tpo) and is highly expressed inearly hematopoietic progenitors and in cells of themegakaryocytic lineage.59 The two most commonMPL mutations, which are located in the cytoplasmicjuxtamembrane portion, are represented by W515L(a tryptophan to leucine substitution) and W515K (atryptophan to lysine substitution; Fig. 1B). Theyhave been detected in 5% to 11% of patients withPMF17,18,60 and in up to 9% of JAK2V617F-negativecases of ET.61,62 Other unusual MPL mutations (egMPLW515S, W5151A, and MPLS505N, initiallydiscovered in association with inherited familialthrombocytosis) have also been reported.63

MPLW515L induced both cytokine-independentgrowth and Tpo hypersensitivity in cell lines, result-ing in constitutively activated JAK-STAT/ERK/Aktsignaling pathways,64 and caused a PMF-like diseasein mice.17 At variance with the JAK2V617F trans-plantation model, the disease induced byMPLW515L was characterized by a rapidly fatalcourse, marked thrombocytosis, leukocytosis, hepa-tosplenomegaly, and bone marrow brosis, all remi-niscent of PMF.17 Interestingly in some patients,multiple MPL mutations or the coexistence withJAK2V617F allele were described.60,62,65

The gene encoding for the receptor of platelet-derived growth factor A (PDGFRA) is involved in atleast four different genetic abnormalities associatedwith eosinophilia.66 The most frequent and bestcharacterized abnormality is due to a karyotypicallyoccult microdeletion at chromosome 4q12, wherePDGFRA is located, resulting in a chimeric FIP1L1-PDGFRA fusion gene (Fig. 1B).10 The latter encodesfor an aberrantly activated tyrosine kinase as theconsequence of disruption of the autoinhibitory ac-tivity encoded by PDGFRA exon 12, where thebreakpoint is located; this constitutively active ty-



rosine kinase drives autonomous eosinophil progen-itor proliferation,67 possesses transforming propertiesin vitro, and induces a myeloproliferative disorderwith extensive eosinophil proliferation when ex-pressed in transplanted mice.68 The fusion gene hasbeen demonstrated at the level of hematopoietic stemcell compartment.69 Also the Beta type of PDGFRhas been reported as being involved in rearrange-ments70 associated with imatinib-responsive eosino-philia.71 The PDGFRB is located at chromosome5q31-32 and may fuse with different partners. One ofthe most common is the ETV6/TEL gene on chro-mosome 12p13, which encodes for a transcriptionfactor with nonredundant roles in normal hemato-poiesis.72 The fusion protein constitutively activatesthe cellular pathways normally associated with PDG-FRB signaling73 and has transforming propertieswhen expressed in cell lines.

A D816V mutation located in the catalytic domainof the tyrosine kinase receptor c-Kit occurs in sys-temic mastocytosis (Fig. 1B).11,74 c-Kit is the recep-tor for stem cell factor, a key cytokine involved in thegeneration and differentiation of mast cells fromprimitive hematopoietic progenitors; it is encoded bykit, located at chromosome 4q12. Additional activat-ing kit mutations other than D816V have also beendescribed in SM, acute leukemia,75 gastrointestinalstromal cell tumors (GIST), and germ cell tumors.76

The D816V and other homologous mutations inducegrowth factor independent growth and cell differen-tiation in mast cell lines through activation ofSTAT5/PI3K/AKT signaling pathways and a phe-notype resembling human SM in murine models.77

Classication of MyeloproliferativeNeoplasmsThe 2008 WHO classication for myeloid neo-plasms, which incorporates novel information de-rived from molecular discoveries in BCR-ABL neg-ative classic myeloproliferative states and clonaleosinophilic disorders, includes ve major entities(Table 2)8 as follows: the Acute Myeloid Leukemia(AML) and the Myelodysplastic Syndromes (MDS)with their different subtypes, whose listing is outsidethe scope of this review; the Myeloproliferative Neo-plasms (MPN); the category of overlapping Myelo-dysplastic/Myeloproliferative Neoplasms (MDS/MPN); and the Myeloid Neoplasms associated with

eosinophilia and specic molecular abnormalities.AML is dened by the presence of either20% blastcells in the bone marrow and/or peripheral blood orcertain characteristic cytogenetic abnormalities.78

The MDSs are recognized and distinguished fromMPN primarily on the basis of the presence of tri-lineage dyshematopoiesis in the absence of monocy-tosis in both bone marrow and peripheral blood.78

A nontrivial formal modication in the 2008WHO classication has been the substitution of theattribute neoplasm for disease. In fact, notwith-standing the analysis of the X chromosome inactiva-tion pattern in informative females and other cyto-genetic and/or molecular ndings that establishedboth classic and nonclassic myeloproliferative dis-orders as being clonal stem cell disorders,79-89 and thending that evolution to AML is part of their naturalhistory,90 the neoplastic nature of these conditions

TABLE 2. The 2008 World Health OrganizationClassication for Myeloid Neoplasms

1. Acute myeloid leukemia (AML) and related precursor neoplasms

2. Myelodysplastic syndromes (MDS)

3. Myeloproliferative neoplasms (MPN)

3.1. Chronic myelogenous leukemia (CML), BCR-ABL1 positive

3.2. Polycythemia vera (PV)

3.3. Essential thrombocythemia (ET)

3.4. Primary myelobrosis (PMF)

3.5. Chronic neutrophilic leukemia (CNL)

3.6. Chronic eosinophilic leukemia, not otherwise classied (CEL-NOS)

3.7. Mastocytosis

3.8. Myeloproliferative neoplasm, unclassiable (MPN-u)

4. Myelodysplastic/Myeloproliferative neoplasms (MDS/MPN)

4.1. Chronic myelomonocytic leukemia (CMML)

4.2. Juvenile myelomonocytic leukemia (JMML)

4.3 Atypical chronic myeloid leukemia, BCR-ABL1 negative

4.4. Myelodysplastic/myeloproliferative neoplasm, unclassiable

4.5. Refractory anemia with ring sideroblasts associated with markedthrombocytosis

5. Myeloid and lymphoid neoplasms with eosinophilia and abnormalitiesof PDGFRA, PDGFRB, or FGFR1

5.1. Myeloid and lymphoid neoplasms associated with PDGFRArearrangement

5.2. Myeloid neoplasms with PDGFRA rearrangement

5.3. Myeloid and lymphoid neoplasms with FGFR1 abnormalities

From Tefferi A and Vardiman JW.136



has been mostly dismissed until recently. This beliefhas most likely represented one of the reasons for thetraditionally poor interest in these neoplasms by can-cer surveillance programs, agencies granting researchsupport, or pharmaceutical companies.

The four classic MPNs (ie, CML, PV, ET, andPMF) should be distinguished from the other non-classic MPNs, which include chronic neutrophilicleukemia (CNL), chronic eosinophilic leukemia-nototherwise specied (CEL-NOS), systemic mastocy-tosis (SM), and unclassiable forms of MPN.91

CML presents very unique characteristics, and it willnot be further discussed herein; recent excellent re-views on molecular and therapeutic issues have beenpublished.92-94

Chronic Neutrophilic LeukemiaCNL is a rare disorder of elderly people characterizedby neutrophilic leukocytosis (greater than 25 109/L)made up of greater than 80% mature granulocytes,splenomegaly, and an absence of the Philadelphia chro-mosome/BCR-ABL fusion gene. Bone marrow biopsyreveals hyperplasia of granulocytic lineage without in-volvement of other series, and there is an absence ofbrosis or myelodysplastic features. Given the potentialfor evolution to acute leukemia or progressive refractoryleukocytosis, allogeneic stem cell transplantation may beappropriate for younger patients.95,96

Chronic Eosinophilic Leukemia andHypereosinophilic SyndromePatients who have a persistent absolute eosinophilcount of at least 1.5 109/L, after exclusion of re-active eosinophilias or other hematologic disorders,suffer from one of the different forms of primaryeosinophilia.97,98 Many patients with nonclonalforms of eosinophilia fall within the category of id-iopathic hypereosinophilia; the possibility of a T-cell mediated eosinophilia, generally via increasedlevels of interleukin-5, can be ruled out with ade-quate studies of T-cell immunophenotyping and T-cell receptor antigen gene rearrangement.99 Con-versely, nding a cytogenetic or molecularabnormality would indicate a clonal, myeloprolifera-tive, eosinophilic disorder.98 Diagnosis of CEL nototherwise (molecularly) specied rests on the dem-onstration of a cytogenetically abnormal proliferation

of eosinophilic precursors with a myeloblast count of5% to 19% in the bone marrow or greater than 2% inperipheral blood, usually accompanied by evidence oforgan damage.97 However, because of intrinsic dif-culties in establishing the presence of a clonal disor-der when the most frequent molecular abnormalitiesassociated with eosinophilia are lacking (see below),it is likely that many forms of CEL-NOS actually fallimproperly within the idiopathic hypereosinophilicsyndrome (HES) category.100 Documentation of tar-get organ damage is necessary for a patient to beconsidered as suffering from HES. Clinical manifes-tations are related to eosinophilic inltration of targettissues and may range from almost asymptomaticdisease to fatal endomyocardial tissue brosis. Bonemarrow biopsy reveals eosinophilia without involve-ment of other cell lines, absence of immature myeloidcells or dysplasia, and a normal number of mast cells.Therapy is based on corticosteroids as rst-line ther-apy, interferon-alpha, or hydroxyurea in refractory orsteroid-dependent patients; some patients may re-spond to imatinib.101 Use of monoclonal antibodiesto interleukin-5 (mepolizumab)102 or CD52 (the re-ceptor for interleukin 2; alemtuzumab)103 has pro-duced appreciable results in refractory cases.66

Mast Cell DiseaseMast cell disease, which is dened by tissue inltra-tion by abnormal mast cells, can be broadly classiedinto cutaneous mastocytosis (CM) and systemic mas-tocytosis (SM); the latter might have an indolent oran aggressive clinical course depending on the ab-sence or presence, respectively, of impaired organfunction.104 Life expectancy is nearly normal in in-dolent forms of SM but is signicantly shortened inaggressive SM. The bone marrow is almost univer-sally involved in SM and is characterized by dense,multifocal aggregates of morphologically and immu-nophenotypically abnormal mast cells, preferentiallyin a perivascular location, and is often accompaniedby increased eosinophils. Serum levels of tryptase aretypically high and represent a clinically useful dis-ease-related marker. By using adequately sensitivemolecular techniques (such as allele-specic poly-merase chain reaction [PCR] amplication of DNA)and mast cell-enriched sources (such as bone marrowaspirate or biopsied lesional material), the kit D816Vmutation is detected in virtually all patients with



SM.105 In addition to its diagnostic value, clinicalrelevance of searching the D816V kit mutation lies inthe almost universally reported refractoriness of mu-tated patients to imatinib.106 Conversely, rare pa-tients with other mutations that are located in thec-Kit juxtamembrane portion may respond to treat-ment with imatinib. Treatment of systemic mastocy-tosis is highly individualized and largely palliative,aiming to prevent or reduce symptoms due to mastcell degranulation or organ inltration. Therapeuticoptions are represented by interferon-alpha or cyto-toxic drugs, such as cladribine, but clinical responsesare limited.107

Myelodysplastic/MyeloproliferativeNeoplasmsMDS/MPN neoplasms are dened by simultaneouspresence of both myelodysplastic and myeloprolifer-ative features that exclude them from being catego-rized as either MDS or MPN alone.108 MDS/MPNneoplasms comprise chronic myelomonocytic leuke-mia (CMML), juvenile MML (of pediatric interestand, thus, not further discussed here), atypical CML,and unclassied MDS/MPN. The clinical and he-matologic presentation of MDS/MPN is pleomor-phic, with cytopenia and dysplasia of any cell lineageeventually becoming associated with elevated leuko-cyte or platelet count. Symptoms may be attributedto cytopenias (anemia, infections, hemorrhage)and/or to myeloproliferation (organomegaly, sys-temic symptoms, cardiovascular events).91 The mo-lecular basis of these disorders is largely unknown,apart from the involvement of ras pathway with mu-tations of RAS109 in CMML and PTPN11110 muta-tions in JMML, or the uncommon presence ofJAK2V617F mutation.111 Therefore, diagnosis relieson a combination of hematological, clinical, and his-tological criteria.

The typical manifestation of CMML is peripheralblood monocytosis greater than 1 106/L and apercentage of monocytes in the white blood cellcount of greater than 10%. Monocytes may or maynot display signs of dysplasia, but the percentage ofimmature monocytes (promonocytes) and mono-blasts in peripheral blood is less than 20%. Bothmonocytic and granulocytic hyperplasia are found in

the bone marrow with a total blast count of less than20%; signs of erythroid and megakaryocytic dysplasiaare variably present. Random cytogenetic abnormal-ities can be discovered in 20% to 40% of these pa-tients. The prognosis is unfavorable with a mediansurvival of only 2-4 years; a major determinant ofsurvival is the percentage of blood and bone marrowblasts.91 Hypomethylating agents like decitabine andazacitidine are now approved for treating CMML.112

Response rate varies from 10% to 30%, depending onthe drug and the schedule used, with best resultsreported for decitabine when a high-dose intensityregimen is used.113-115 Treatment is well toleratedwith relatively few nonmyelosuppressive complica-tions and has become a valuable therapeutic optionfor a disease where just a few years ago the standardof care was merely supportive.

Atypical CML is a rare, aggressive disorder with amedian survival of 1-2 years and usually affects el-derly patients. It presents features typical of classicCML, but unlike classic CML, it is BCR/ABL neg-ative and displays manifest signs of dysgranulopoiesiswith nuclear aberrations and cytoplasmic hypogranu-lation.108 JAK2V617F mutation is absent.116 Thebone marrow is hypercellular; dysplasia of myeloidlineage with less than 20% blasts is a constant feature,whereas other lineages are variably involved. Thedisease terminates in AML in up to 40% of thesepatients.

Refractory anemia with ring sideroblasts andthrombocytosis (RARS-T) is a rare syndrome char-acterized by anemia with dyserythropoiesis and ringsideroblasts in the bone marrow, associated withthrombocytosis and increased number of largemegakaryocytes. These morphologic features ofmegakaryocytes are distinct from the appearance typ-ically associated with the 5q- abnormality.117 A highproportion of patients have the JAK2V617F muta-tion,118 although a few harbor the MPL mutation.119

RARS-T is a disease with a relatively good progno-sis,120 and it shares many aspects with classic MPN.

Finally, when myelodysplastic and myeloprolifer-ative features simultaneously present in bone marrowaspirate do not t into any of the previous categories,and after any known molecular or cytogenetic abnor-mality is excluded, the disorder is dened as MDS/MPN unclassiable, with a comment describing theatypical features.108



Myeloid Neoplasms Associated withEosinophilia and Specic MolecularAbnormalitiesBoth the alpha (PDGFRA) and beta (PDGFRB)types of platelet-derived growth factor receptor(PDGFR) genes may be involved in genetic abnor-malities associated with eosinophilia.121 Eosinophiliaassociated with FIP1L1-PDGFRA rearrangementhas a strikingly male predominance. In addition to anexpanded eosinophilic lineage, the bone marrow of-ten contains an increased number of mast cells that,together with ndings of raised serum levels oftryptase, sometimes make problematic the differen-tial diagnosis with SM.122 However, at variance withkit D816V-mutated forms of SM, presence of theFIP1L1-PDGFRA mutation reliably predicts hema-tologic and molecular remission when imatinib isused at doses lower than those used for CML (100mg daily is generally efcacious).123 The rate of com-plete molecular response may be as high as 95%, anda prospective multicenter study showed it to be stableand durable during a median follow-up of 25 monthsbut to be dependent on treatment continuation. Inthree patients who discontinued imatinib, molecularnegativity was lost and then regained after imatinibwas resumed.124 A denitely lower proportion ofpatients with imatinib-responsive eosinophilia haverearrangements involving the PDGFRB gene. Fi-nally, translocations involving the broblast growthfactor receptor-1 gene (FGFR1), which is located atchromosome 8p11, and several different gene part-ners are at the basis of the 8p11 myeloproliferativesyndrome.125 This is also called stem cell leukemia/lymphoma syndrome because of the clinical pheno-type that is characterized by features of both lym-phoma and eosinophilic myeloproliferation. Thedisease results from constitutive activation of the ty-rosine kinase domain of FGFR1 after its juxtapositionwith any partner gene. Prognosis is very poor with mostpatients progressing to overt AML or lymphoblasticlymphoma with 1-2 years of diagnosis.126

The Classic MyeloproliferativeNeoplasmsAmong classic MPNs, PV and ET are relativelyindolent disorders,127 resulting in a modest reductionof lifespan compared with a control population; how-

ever, most patients ultimately suffer from one ormore severe and potentially fatal complications di-rectly attributable to the disease. Conversely, PMFhas a severe course in most cases, and survival issignicantly affected. The three clinical entities shareseveral common features,6 such as their origin in amultipotent hematopoietic stem cell, a relatively nor-mal cellular maturation, a striking overlap in clinicalpresentation (apart from PMF, which has its ownpeculiar manifestations), and in cases of PV and ET,the propensity to evolve into post-polycythemic orpost-thrombocythemic myelobrosis (or less fre-quently each into the other), and the possibility totransform into AML.90

EpidemiologyClassic MPNs are among the most frequent hema-tologic neoplasms, usually affecting the adult elderlypopulation; however, they can also be found in chil-dren, and in this instance, they raise specic diag-nostic and management issues128,129 that are beyondthe scope of this review. A recent study,130 based onthe North American Association of Central CancerRegistries (NAACCR) encompassing 82% of totalUS population, reported an average 2001-2003 an-nual age-adjusted incidence rate of 2.1 per 100,000and estimated that there were 6,328 new cases in thetotal US population in 2004. Furthermore, becauseof their relatively smooth clinical course, it is likelythat many classic MPN cases actually go undetectedor are not reported to registries. Advanced age, malesex, and white race were identied as risk factors.Among individuals aged 80 years or older, the rate wasas high as 13.3 per 100,000. Familial clustering of thesedisorders is known, and even before the discovery ofJAK2V617F mutation, 131 this observation led to asuggestion of predisposition allele(s).132 This hypoth-esis gained substantial support from a large studyrecently completed in Sweden.133 Relatives of pa-tients with MPN had a 5.7 relative risk (RR) ofhaving PV, an RR 7.4 for ET, and an RR of 7.5 forunclassied forms of MPN, together with a border-line increased RR of CML. The higher risk observedamong siblings would suggest a model of recessiveinheritance, although whether presentation of diseaseoccurs at an even younger age than in parents isdebatable.133,134 Accordingly, thorough investigationof family history is mandatory in the initial workup ofpatients with classic MPN, and appropriate counsel-



ing should be provided. The coexistence of differentclinical entities and of JAK2V617F-positive andJAK2V617F-negative diseases in the same family isnoteworthy.131,134,135

DiagnosisBecause of similarities with reactive forms character-ized by an increased count of mature peripheral bloodcells on one side, and the signicant phenotypicoverlapping among them on the other, diagnosis ofdifferent MPNs has traditionally been challenging;the availability of the new molecular markers is ex-pected to facilitate diagnosis (Table 3). As a matter offact, molecular genotyping is integral to the 2008WHO diagnostic criteria,136 and tests for JAK2 orMPL mutation already have become a standard toolin the diagnostic work up of MPN (Fig. 2).137 In fact,detection of one of these mutations unequivocallyestablishes by itself the presence of a clonal MPN andrules out the possibility of reactive erythrocytosis,thrombocytosis, or myelobrosis. Unfortunately,they are of no help in distinguishing among thedifferent forms of MPNs, although JAK2 exon12

FIGURE 2. Rationale for using JAK2V617F genotyping in the diagnosticwork-up of suspected MPN. See text for details.

TABLE 3. 2008 WHO Diagnostic Criteria for classic MPN

CRITERIA POLYCYTHEMIA VERA ESSENTIAL THROMBOCYTHEMIA PRIMARY MYELOFIBROSIS

Major criteria 1. Hgb 18.5 g/dL (men) or 16.5 g/dL(women)

or Hgb or Hct 99th percentile ofreference range for age, sex, oraltitude of residence

or Hgb 17 g/dL (men) or 15 g/dL(women) if associated with adocumented and sustained increaseof 2 g/dL from baseline thatcannot be attributed to correction ofiron deciency

or elevated red cell mass 25% abovemean normal predicted value

2. Presence of JAK2V617F or similarmutation

1. Sustained platelet count 450 x 109/L2. BM showing proliferation mainly of themegakaryocytic lineage with increasednumbers of enlarged, maturemegakaryocytes. No signicant increase orleft-shift of neutrophil granulopoiesis orerythropoiesis3. Not meeting the WHO criteria for PV,PMF, CML, or MDS or other myeloidneoplasm4. Demonstration of JAK2V617F or otherclonal marker

or no evidence of reactive thrombocytosis

1. Megakaryocyte proliferation and atypia*accompanied by either reticulin and/or collagenbrosis

or In the absence of reticulin brosis, themegakaryocyte changes must be accompaniedby increased marrow cellularity, granulocyticproliferation and often decreasederythropoiesis (ie, pre-brotic cellular-phasedisease)

2. Does not meet WHO criteria for CML, PV, MDS,or other myeloid neoplasm3. Demonstration of JAK2V617F or other clonalmarker

or no evidence of reactive marrow brosis

Minor criteria 1. BM showing hypercellularlity for ageand trilineage growth (panmyelosis)

1. Leukoerythroblastosis

2. Subnormal serum Epo level 2. Increased serum LDH

3. EEC growth 3. Anemia

4. Palpable splenomegaly

Diagnosticcombinations

Both major criteria 1 minor criterionor rst major criterion 2 minor

criteria

All 4 criteria must be met All 3 major criteria 2 minor criteria

WHO indicates World Health Organization; MPN, myeloproliferative neoplasm; CML, BCR-ABL1 chronic myelogenous leukemia; PV, polycythemia vera; PMF, primarymyelobrosis; MDS, myelodysplastic syndrome; BM, bone marrow biopsy specimen; Epo, erythropoietin; EEC, endogenous erythroid colonies; LDH, lactate dehydro-genase.*Small to large megakaryocytes with an aberrant nuclear/cytoplasmic ratio and hyperchromatic, bulbous, or irregularly folded nuclei and dense clustering.



mutations have not yet been reported outside PV,and no patient with PV has been found to harbor anMPL mutation.

In patients with evidence of increased red cellmass, according to WHO criteria,136 demonstrationof JAK2V617F mutation allows a diagnosis in greaterthan 95% of cases, as less than 2% of PV patientsharbor JAK2 exon 12 abnormalities.16 It is debatedwhether a diagnosis of PV can still be tenable in theabsence of JAK2 mutation.138,139

The compelling criterion for a diagnosis of ET isa sustained platelet count of greater than 450 109/L. Notably, this value is lower than the one originallyused by the 2001 WHO classication system(600 109/L),5 because the latter might have led toinadvertently overlooking classic ET cases with alower platelet count.140 This assumption is supportedby the discovery of the JAK2V617F mutation in somesubjects who have a platelet count lower than600 109/L.47 Diagnosis of ET requires exclusionof reactive thrombocytosis141,142 as well as of otherMPNs that present with thrombocytosis. In partic-ular, exclusion of CML with FISH or PCR analysisfor BCR-ABL rearrangement is mandatory. Positivityfor JAK2V617F or MPL mutation cumulatively ac-count for 60% to 70% of ET cases. Therefore, theassessment of bone marrow morphology remains keyto the diagnosis of ET; bone marrow cellularity isnormal or slightly increased, with abundance of large,mature-appearing megakaryocytes devoid of mor-phological abnormalities and generally dispersedthroughout the biopsy. This appearance is distinctfrom both the panmyelosis typical of PV or thepredominant granulocytic hyperplasia with highly bi-zarre megakaryocytes, often found in abnormallytight clusters, with aberrant nuclear to cytoplasmicratio and hyperchromatic, bulbous, or irregularlyfolded nuclei that are found in PMF, even in initialstages without overt brosis.137,143

Bone marrow histology is required for the diagno-sis of PMF. Although advanced reticulin or colla-genic brosis is typically associated with classic stagesof PMF, some degree of reticulin brosis can befound as well as in PV, or more occasionally in ET.Therefore, brosis by itself is not synonym for PMF,and diagnosis of PMF can be made even in theabsence of overt brosis.8 Also the leukoerythroblas-tic features of blood smears, with immature myeloidprecursors, nucleated red cells, and abnormally

shaped erythrocytes (tear-drop cells), is very charac-teristic, but not diagnostic, of PMF. CML should beruled out through BCR-ABL rearrangement analysis,while nding a positive JAK2V617F or MPL muta-tion allows exclusion of reactive forms of myelobro-sis (such as in infectious or inammatory processes,metastatic cancer, and lymphoid disorders). Somecytogenetic abnormalities, such as del(13)(q12;q22),are frequently encountered and may be diagnosticallyspecic in this context.144 Anemia, palpable spleno-megaly, and raised lactate dehydrogenase levels areadditional diagnostic criteria.8

Clinical Course and Risk StraticationThrombosis, hemorrhage, evolution to post-polycy-themic or post-thrombocythemic myelobrosis, andAML transformation represent the most clinicallyrelevant issues in the course of classic MPN.145-147

Most thrombotic events occur at or in the two yearsbefore diagnosis.148 However, epidemiologic infer-ence from the European Collaboration on Low-doseAspirin in Polycythemia (ECLAP) study146,149 andthe UK Medical Research Council Primary Throm-bocythemia-1 (MRC PT-1) study150 suggested thatthe cumulative rate of thrombosis during the diseasecourse ranged from 2.5% to 5.0% and from 1.9% to3% per patient-year in PV and ET, respectively,depending on whether the patient was in a low-riskor high-risk category.146,150 In a large retrospectivestudy of PV or ET patients who had suffered from aprevious cardiovascular event, the calculated recur-rence rate was 5.6% patient-year with a cumulativeprobability of 49.9% at 10 years.151 Arterial throm-bosis accounts for 60% to 70% of all cardiovascularevents and includes acute myocardial infarction, isch-emic stroke, and peripheral arterial occlusion. Eventsinvolving the venous system, more common amongPV patients, are represented by lower extremity deepvenous thrombosis, pulmonary embolism, andsplanchnic vein thromboses (SVT, which includesportal vein thrombosis, mesenteric thrombosis, andthrombosis of the hepatic veins causing Budd-Chiarisyndrome). The prevalence of SVT is unusually highamong MPN patients152; however, diagnosis is oftencomplicated by the hemodilution resulting from hy-persplenism that makes blood cell counts unreliable,in particular as concerns evidence of increased red cellmass necessary for the diagnosis of PV.153 Recentdata indicate that at least 40% of patients with SVT



not attributable to other causes actually harbor theJAK2V617F mutation; therefore, JAK2V617F geno-typing represents a rst-line test for these condi-tions.154,155 Occasional SVT patients harboring MPLmutation have also been reported.156 Conversely, in-volvement of the microcirculatory system is moretypically associated with ET and manifests as eryth-romelalgia (a rare disorder characterized by burningpain, warmth, and redness of the extremities due toarteriolar brosis and occlusion with plateletthrombi, typically aspirin-sensitive),157 transientischemic attacks, visual or hearing transitory defects,recurrent headache, and peripheral paresthesia; how-ever, because of the lack of objective diagnostic cri-teria, true incidence of microvessel disturbances isdifcult to assess.158 Pathogenesis of thrombosis inclassic MPNs is multifactorial; rheologic abnormali-ties due to increased red cell mass in PV, abnormalfunction of platelets and their enhanced interactionwith leukocytes and endothelial cells, are all possiblecontributing factors159; however, neither thrombocy-tosis nor increased hematocrit (at least until 52%) areclearly associated with occurrence of thrombosis.160

Mortality rate is age-dependently increased in PV,being 1.6-fold and 3.3-fold higher than in the refer-ence population in patients younger or older than 50years, respectively.161 Conversely, survival of ET pa-tients is reduced by about 2-fold compared with thegeneral population starting from the rst decade afterdiagnosis.162 Major causes of shortened survival inPV or ET are represented by thrombotic events andtransformation to myelobrosis or AML, which ac-count for 41% and 13% of total deaths among 1,638PV patients that were included in the observationalarm of the ECLAP study.146,163 An age of greaterthan 60 years and leukocytosis were incorporated in apredictive model for survival in ET that discrimi-nated groups of patients with median survivals of 25,17, and 10 years, respectively.162 Therefore, becauseof the nding that thrombosis represents the mostcommon event that complicates the courses of PVand ET, and eventually is the leading cause of death,it seems appropriate to use this clinical endpoint asthe criterion for stratifying patients according to theirrisk.164 Older age (greater than 60 years) and a pre-vious history of thrombosis are standard risk factorsfor thrombosis in both PV and ET (Table 4), whichhave been validated in several studies.146,148,165 In thepresence of either of these, a patient is at high-risk,

whereas when neither of these is present, the diseaseis low-risk. The role of generic cardiovascular riskfactors, such as hypertension, diabetes, hyperlipid-emia, smoking, or genetic alterations of hemostaticfactors, is still controversial; however, patients whopresent with any of these abnormalities are pruden-tially considered to belong to an intermediate-riskcategory,158,166 and both specic medical interventionand correction of life style issues, particularly smok-ing, should be aggressively pursued. Recent studieshave demonstrated that leukocytosis is an additionalindependent risk factor for thrombosis,162,167,168 par-ticularly for acute myocardial infarction in PV. Fur-thermore, low-risk ET patients could be separatedinto two categories with a respective overall preva-lence of thrombosis of 55% and 20% depending onthe presence, or not, of an absolute leukocyte countgreater than 8.7 109/L.169 Finally, there is alsoevidence that JAK2V617F mutated status inET,47,170,171 and a high V617F allelic burden in bothET47,172 and PV172-173 are associated with increasedrisk of thrombosis. Therefore, both leukocytosis andJAK2V617F mutated status represent novel, power-ful, disease-associated, risk factors; however, beforethey are included in current risk stratication criteriaoutlined in Table 4, they need validation in prospec-tive studies.

Life expectancy in PMF is 31% lower than in anage-matched and sex-matched population, with amedian survival of 5 years, although younger patientsmay experience longer survival.161,174,175 Major causesof death are represented by the sequelae of portalhypertension or hepatic-splenoportal thrombosis,thromboses in various anatomic sites, heart failuredue to splenic pooling, infections, pulmonary hyper-tension, bleeding caused by thrombocytopenia or he-mostatic defects, and transformation to AML.147

Prognostic staging systems for PMF have been de-veloped that allow separation of patients with low-

TABLE 4. Risk-Stratication of Patients with PolycythemiaVera or Essential Thrombocythemia

RISK CATEGORY

AGE >60 YEARSOR HISTORY OFTHROMBOSIS

GENERIC CARDIOVASCULARRISK FACTORS

Low No No

Intermediate No Yes

High Yes Irrelevant



risk and high-risk disease associated with signi-cantly different survival times (Table 5). The mostused Lille score includes anemia and abnormal leuko-cyte count as variables and effectively distinguishes pa-tients with survival times that range from 1-8 years.176

Stratication according to risk is of particular impor-tance in younger patients who may potentially exploitthe curative efcacy of allogeneic hematopoietic stemcell transplantation (HSCT). In this regard, Cer-vantes175 and Mayo ad hoc scoring systems for pa-tients aged younger than 55 years or 60 years have beendeveloped177 and represent useful instruments to aidboth physician and patient to make the most appropri-ate therapeutic decision (Table 5). Presence of aJAK2V617F mutated state independently predictedleukemic transformation in a longitudinal prospectiveseries of PMF patients,178 whereas presence ofMPLW5151L/K mutation was associated with moresevere anemia.60 However, because of conicting resultsreported in similar studies,179 these markers need fur-ther validation before being operationally incorporatedinto prognostic systems.

Transformation to post-polycythemic or post-thrombocytemic myelobrosis represents the natural

evolution of PV and ET, occurring late in the clinicalcourse. The estimated rate is about 5% after 15 yearsfrom diagnosis of PV,146 whereas data are scanty in ET.Criteria for the diagnosis of evolution to myelobrosishave recently been proposed by the InternationalWorking Group for Myelobrosis Research and Treat-ment (IWG-MRT; Table 6).180 Survival is probablyshortened by the development of myelobrosis, andmay be predicted by hemoglobin level and platelet andleukocyte counts according to a dynamic prognosticmodel recently developed in PV patients.183

Evolution to AML occurred in 1.3% of PV patientsincluded in ECLAP study, at a median time of 8.4years after diagnosis163; however, because of the shortfollow up, a precise estimate cannot be made, andinformation about ET is not yet available. Survival timeis dismal, less than 6 months, although recipients ofallogeneic HSCT may experience longer remission.184

Advanced age, elevated leukocyte count, and longerdisease duration were factors associated with increasedrisk of leukemic transformation.163 An increased risk ofAML was reported in patients who were treated withradioactive phosphorus or chlorambucil in the PVSGtrial.185 In addition, sequential or combined use of more

TABLE 5. Prognostic Scoring Systems Used for Risk Assessment in Patients with PMF

PROGNOSTIC SCORING SYSTEM PROGNOSTIC FACTORS

NO. OF PROGNOSTIC FACTORS BY RISKCATEGORY

NO. OF MONTHS OF SURVIVAL BY RISKCATEGORY

LOW INTERMEDIATE HIGH LOW INTERMEDIATE HIGH

All patients

Lille Hb 10 g/dLWBC 4 or30x109/L

0 1 2 93 26 13

Cervantes Hb 10 g/dLPB Blasts 1%Constitutionalsymptoms

0-1 2-3 99 21

Mayo Hb 10 g/dLWBC 4 or30x109/LPlt 100x109/LMonocytes 1x109/L

0 1 2 173 61 26

Younger patients

Cervantes, aged 55 y Hb 10 g/dLPB Blasts 1%Constitut. symptoms

0-1 2-3 176 33

Dingli, aged 60 y Hb 10 g/dLWBC 4 or30x109/LPlt 100x109/L

0 1 2-3 155 69 24

Constitutional symptoms included unexplained fever, night sweats, or weight loss of greater than 10% of baseline value in the last 6 months.

PMF indicates primary myelobrosis; Hb, hemoglobin; WBC, white blood cell count; PB Blasts, percentage of blasts in peripheral blood smears; Plt, platelet count.



than one chemotherapeutic agent, including hydroxy-urea (HU), signicantly increased the rate of evolutionto AML in PV patients in the observational arm ofECLAP study.186

Management of Classic MPNOver the years, there has been a shortage of clinicalstudies specically devoted to classic MPN. Mostavailable information derives from a limited numberof randomized clinical trials performed within na-tional or international collaborative groups that in-clude the Polycythemia Vera Study Group(PVSG),165,186 the ECLAP study,146,149 the Ber-gamo trial,187 and the PT-1 trial150; however, theinformation they have provided represents the foun-dation for current treatment indications189-192 as well

as the basis for future studies(Table 7). Standardized criteriafor assessing clinical and hema-tologic responses in PMF havebeen published193,194 and will beof particular relevance for evalu-ation of novel molecularly targetdrugs. Conversely, similar crite-ria for patients with PV or ETare still lacking.

Cytoreductive Therapy in PVand ETTreatment of patients with PVor ET should adhere to the stan-dard risk stratication outlinedabove (Table 4). Phlebotomy isthe cornerstone of treatment inlow-risk patients with PV,aimed at reaching and maintain-ing a target hematocrit of lessthan 45% in men and less than42% in women, according tostandard recommendations.165

The ultimate goal of this prac-tice is to limit availability of ironto erythropoiesis, but often itwill cause symptoms due to se-vere and prolonged iron de-ciency; fatigue is being recog-nized as one major burden forquality of life in patients withPV.127 In fact, there is wide vari-

ability in opinions and attitudes among US andnon-US physicians concerning the optimal hemato-crit target to be attained with phlebotomies.195 Con-versely, high-risk patients should receive myelosup-pressive therapy, eventually in association withphlebotomy, and hydroxyurea (HU) is the drug ofchoice (Table 7). HU is an antimetabolite thatprevents synthesis of DNA. It is also approved forthe treatment of sickle cell anemia because of itscapacity to reactivate synthesis of hemoglobin F,resulting in a signicant decrease of occlusive andhemolytic events.196 Superiority of HU comparedwith phlebotomy was suggested in a comparativeanalysis of the PVSG in the 1980s,185 but norandomized trial to address this issue has yet beenundertaken.

TABLE 6. Criteria for Establishing the Diagnosis of Evolution to Post-polycythemicor Post-thrombocythemic Myelobrosis According to IWG-MRT Criteria180

CRITERIA FOR POST-POLYCYTHEMIC MYELOFIBROSIS

Required criteria

1. Documentation of a previous diagnosis of polycythemia vera as dened by WHO criteria136

2. Bone marrow brosis grade 2-3 (according to the European classication181) or grade 3-4 (accordingto standard classication182)

Additional criteria

1. Anemia (below the reference range for appropriate age, sex, and altitude considerations) or sustainedloss of either phlebotomy (in the absence of cytoreductive therapy) or cytoreductive treatmentrequirement for erythrocytosis

2. A leucoerythroblastic peripheral blood picture

3. Increasing splenomegaly of 5 cm (distance of the tip of the spleen from the left costal margin) orthe appearance of a newly palpable splenomegaly

4. Development of 1 of 3 constitutional symptoms: 10% weight loss in 6 months, night sweats,unexplained fever (37.5C)

CRITERIA FOR POST-THROMBOCYTHEMIC MYELOFIBROSIS

Required criteria

1. Documentation of a previous diagnosis of essential thrombocythemia as dened by WHO criteria136

2. Bone marrow brosis grade 2-3 (according to the European classication181) or grade 3-4 (accordingto standard classication182)

Additional criteria

1. Anemia (below the reference range for appropriate age, sex, and altitude consideration) and a 20g/L decrease from baseline hemoglobin level

2. A leucoerythroblastic peripheral blood picture.

3. Increasing splenomegaly of 5 cm (distance of the tip of the spleen from the left costal margin) orthe appearance of a newly palpable splenomegaly

4. Increased LDH (above reference level)

5. Development of 1 of 3 constitutional symptoms: 10% weight loss in 6 months, night sweats,unexplained fever (37.5C)

Diagnosis is made on the basis of meeting all required criteria plus two additional criteria.

WHO indicates World Health Organization; LDH, lactate dehydrogenase.



The use of low-dose aspirin in PV was exploited inthe ECLAP study that randomized 518 low-riskpatients in a double-blind, placebo-controlled trial.149

The primary study endpoint (cardiovascular death,nonfatal myocardial infarction, nonfatal stroke, andmajor venous thromboembolism) was signicantlylowered by aspirin (RR, 0.40; 95% condence inter-val [CI] 0.18 to 0.91; P .02), with only a small,nonsignicant increase of major hemorrhage (RR,1.62; 95% CI, 0.27 to 9.71; P .60); total andcardiovascular mortality were also reduced by 46%and 59%, respectively. Therefore, in the absence of ahistory of major bleeding, allergy to the drug orsevere asthma, or gastric intolerance, low-dose aspi-rin (100 mg daily) is recommended regardless of riskcategory for all patients with PV.149

Low-risk patients with asymptomatic ET do notneed therapy, although high-risk patients have thesame indications for the use of HU as patients withPV.197 In the Bergamo trial, which randomized 114high-risk patients to HU versus no treatment, thepercentage of patients who developed thrombosisdecreased from 24% to 3.6%.188 HU was also superiorto anagrelide, a nonmyelosuppressive platelet-lower-ing drug, in preventing arterial thrombosis in therandomized MRC-PT-1 trial, which included 809high-risk patients, although venous thrombosis wasreduced in the anagrelide arm.54 Interestingly,JAK2V617F-positive patients had a better responseand required lower doses of HU to control throm-bocytosis compared with patients who did not havethis mutation.54 The target level at which the plateletcount should be maintained with therapy in high-riskpatients is currently set at 400-450 109/L, but thisis not based on evidence.197,198 Also, there is littlerationale for the use of cytoreductive treatment toreduce extreme thrombocytosis (platelet countgreater than 1,000 109/L) in an otherwise low-

risk, asymptomatic patient.199

Unlike PV, the safety and ef-cacy of low-dose aspirin use forET has not formally beenproven, but most patients in in-termediate-risk or high-risk cat-egories are currently advised touse the drug. Higher doses, upto 500 mg daily, may be requiredfor acute symptoms because ofmicrovascular disturbances, in

particular erythromelalgia. Conversely, extremethrombocytosis is considered a contraindication foraspirin use because of a possibly increased bleedingtendency due to acquired Von Willebrand dis-ease.158,200-202

There has been some debate about potential leuke-mogenicity of HU, but although current evidence doesnot attribute to the drug a denite risk in this regard, itis appropriate to reserve HU use for patients at high riskof developing complications and in whom benets ex-pected from treatment overcome potential unwantedeffects. Actually, transformation to AML is consideredpart of the natural history of MPN.186

Noncytotoxic Drugs for PV and ETInterferon-alpha (IFN-), a nonleukemogenic agent,has multiple potential activities against hematopoieticprogenitor cell proliferation and differentiation whichmay justify its use in the youngest of patients with PVand ET. However, tolerance is often poor because ofacute and chronic side effects that cause discontinuationof the drug in one-third of patients. IFN- has beenshown to effectively reduce the hematocrit or plateletcount to a target level in the majority of cases,203,204 andno thrombohemorragic events were recorded among 55patients with PV who were followed for a median of 13years.205 Progressive decrease of JAK2V617F burdenhas been suggested in one study,206 whereas changeswere minimal in another study.207 Notably, a recent,phase 2, multicenter study that used pegylated-IFN-in 40 patients with PV reported complete molecularremission in 7 of these patients.208 Finally, becauseIFN- is not teratogenic and does not cross the pla-centa, it is recommended whenever there is the need forcytoreduction during pregnancy, according to currentguidelines.209

Anagrelide has widely been used to control plateletcount in patients with ET in all risk categories210; the

TABLE 7. Risk-Oriented Therapy in Polycythemia Vera (PV) and EssentialThrombocythemia (ET)

RISK CATEGORY RISK FACTORS PV ET

Low Age 60 y and no priorcardiovascular event Phlebotomies plus low-dose

aspirin

Nil, or low-dose aspirin(no consensus)

Intermediate Generic cardiovascularrisk factors

Low-dose aspirin (noconsensus)

High Age 60 y and/or priorcardiovascular event

MyelosuppressionPhlebotomies Myelosuppression

Low-dose aspirin Low-dose aspirin



majority of patients achieved adequate control ofthrombocytosis, although cardiovascular side effects(mainly palpitations and headache, less frequentlycongestive heart failure) may require early discontin-uation of treatment.211,212 The drug is considereddevoid of any leukemogenic potential, but it shouldnot be prescribed during pregnancy. On the basis ofresults of the PT-1 trial, anagrelide is not recom-mended in high-risk patients as an alternative to HU,and its rationale in otherwise low-risk or intermedi-ate-risk patients should be carefully evaluated case bycase.213 However, according to recently publishedguidelines,214 anagrelide could be successfully used inplatelet count control in patients with PV or ET whoare refractory and/or resistant, or who show poortolerance, or who develop side effects to HU.

Management of PMFThe only approach that has resulted in a prolongationof survival time in PMF and has the potential to becurative is allogeneic HSCT.215 At present, it should bereserved for patients with high-risk disease after carefulclinical evaluation and thorough patient counseling,particularly considering the option of inclusion in trialswith innovative drugs. Bothmyeloablative and reduced-intensity conditioning regimens have been used, withsimilar efcacy in terms of survival (3-year event-freesurvival in the range of 50% to 60%) but lower mortalityrate with the use of the latter in older patients.216

Therefore, a myeloablative strategy may be consideredas the most appropriate for younger patients, whereasthe reduced-intensity regimen would be the best forolder patients. Furthermore, in patients who relapseafter HSCT, a graft-versus-myelobrosis effect couldbe demonstrated after donor-lymphocyte infusion witha remarkable reduction of bone marrow brosis.217,218

Factors that have been reported as having a favorableimpact on overall survival after HSCT include a con-ditioning regimen with busulfan/cyclophosphamide,younger age, high platelet count, low comorbidity in-dex, low risk according to the Dupriez score, normalkaryotype, hemoglobin of greater than 100 g/L, absenceof circulating blasts, and absence of osteosclerosis.219-221

The usefulness of pretransplant splenectomy still re-mains controversial.215,222,223

Given that a conventional drug therapy does notsignicantly modify disease course and is largely in-effective, it is reserved for patients who presenteither with symptomatic anemia or splenomegaly.

Androgens,224 prednisone,224 erythropoiesis-stimulat-ing agents,225-227 and danazol228,229 are all variably usedwith measurable effect in a few patients. Low-dosethalidomide in combination with prednisone improvesanemia or thrombocytopenia in 30% to 50% of cas-es.230-234 Lenalidomide, a thalidomide analog, has pro-duced excellent and durable responses in the relativelyinfrequent PMF patients who have the del(5q) abnor-mality,235 and it can be recommended as rst-line ther-apy in this patient subset. When there is the need tocontrol excessive myeloproliferation, ie, leukocytosis,thrombocytosis, or progressive splenomegaly, HU is thecurrent drug of choice.236 Several other drugs, includingbusulfan,237 melphalan,238 and 2-chlorodeoxyade-nosine,239 have been used particularly in HU-refractorypatients, but results are generally dismal. Splenectomyhas a role for alleviating mechanical symptoms due toextreme splenomegaly and can also ameliorate anemiain approximately 25% of transfusion-dependent pa-tients.240 However, splenectomy in PMF bears an ap-proximately 10% procedure-related mortality, and itshould be performed by experienced surgeons. Further-more, up to 25% of patients present with acceleratedhepatomegaly and extreme thrombocytosis after sple-nectomy, and these patients require further cytoreduc-tion.240,241 Splenic irradiation is reserved for patientswho cannot undergo splenectomy for any reason, butthe efcacy of this therapy is poor, and subsequentcytopenias are often severe. Conversely, radiation ther-apy has a dened role in the treatment of nonhepato-splenic extramedullary hematopoiesis, such as in casesof spinal cord compression by foci of eterotopic hema-topoiesis.242-245

Prospect for Molecularly Targeted Therapy inClassic MPNThe involvement of JAK-STAT pathways in most pa-tients who have classic MPN and harbor mutations inJAK2 or MPL and the experimental evidence that sug-gests that the same signaling abnormalities may be atthe basis of mutation-negative patients20 are behindactive efforts to develop anti-JAK2 drugs. Many mol-ecules have undergone preclinical testing, in vitro andalso in vivo, and some have already been introduced intoclinical trials.246-252 A very incomplete list of moleculesthat may or may not have selective anti-JAK2 activity isreported in Table 8. Concerning selective JAK2 inhib-itors, we have listed only those that are already inclinical trials or whose activity has been demonstrated in



JAK2V617F-mutated murine models (TG101348).248

Among these, INCB018424, XL019, CEP-701, andTG101348 are currently undergoing clinical trials inpatients with advanced stages of PMF, post-PV/ETmyelobrosis, PV, and JAK2V617F-positive ET.253-256

Preliminary results have been encouraging in terms ofactivity against splenomegaly and constitutional symp-toms,253 with minimal toxicity. Although the numberof patients treated until now is less than 100 with anysingle drug and, thus, prevents us from making anydenitive comment, the hope that this molecularly tar-

geted approach may nally result in improving qualityof life and possibly the chance of cure for patients withclassic MPN is enormous.

Patient ResourcesDuring the last few years, we have witnessed a re-newed interest in the MPN eld among scienticcommunities and pharmaceutical companies; at thesame time, the patient community is growing inawareness and strength. There are several focusedresources for patient information and support thatinclude the Myeloproliferative Disorders ResearchConsortium (MPD-RC, an international researchconsortium funded by the National Cancer Institute;http://www.mpd-rc.org), theMyeloproliferative Dis-orders Foundation (committed to promoting focusedresearch and international expert cooperation andalso devoted to patient education and support; http://www.mpdinfo.org), the Mastocytosis Society (http://www.tmsforacure.org), and several online supportgroups (such as http://www.acor.org; http://www.mpdsupport.org). Among non-US resources arethe Myeloproliferative Disorders Australia (MPD-Oz; http://www.mpd-oz.org), the Italian Mielo-brosi Insieme (for patients with PMF; http://www.myelobrosis.net), and the Gruppo Italiano perle Malattie Ematologiche Maligne dellAdulto-GIMEMA (http://www.gimema.org).

REFERENCES1. Dameshek W. Some speculations on the

myeloproliferative syndromes. Blood1951;6:372375.

2. Mesa RA, Verstovsek S, Cervantes F,et al. Primary myelobrosis (PMF),post polycythemia vera myelobrosis(post-PV MF), post essential thrombocy-themia myelobrosis (post-ET MF),blast phase PMF (PMF-BP): Consensuson terminology by the internationalworking group for myelobrosis re-search and treatment (IWG-MRT). LeukRes 2007;31:737740.

3. Nowell PC, Hungerford DA. Chromosomestudies on normal and leukemic humanleukocytes. J Natl Cancer Inst 1960;25:85109.

4. Tefferi A. The history of myeloprolifera-tive disorders: before and after Dameshek.Leukemia 2008;22:313.

5. Jaffe ES, Harris NL, Stein H, Vardiman JW.World Health Organization Classication ofTumors of Hematopoietic and LymphoidTissues. Lyon: IARCPress; 2001.

6. Spivak JL. Diagnosis of themyeloprolifera-tive disorders: resolving phenotypic mim-icry. Semin Hematol 2003;40:15.

7. Kaushansky K. The chronic myeloprolif-erative disorders and mutation of JAK2:Damesheks 54 year old speculation comesof age. Best Pract Res Clin Haematol2007;20:512.

8. Tefferi A, Thiele J, Orazi A, et al. Proposalsand rationale for revision of the WorldHealth Organization diagnostic criteria forpolycythemia vera, essential thrombocy-themia, and primarymyelobrosis: recom-mendations from an ad hoc internationalexpert panel. Blood 2007;110:10921097.

9. Rowley JD. Letter: A new consistent chro-mosomal abnormality in chronic myelog-enous leukaemia identied by quinacrineuorescence and Giemsa staining. Nature1973;243:290293.

10. Cools J, DeAngelo DJ, Gotlib J, et al. Atyrosine kinase created by fusion of thePDGFRA and FIP1L1 genes as a therapeu-tic target of imatinib in idiopathic hyper-eosinophilic syndrome. N Engl J Med2003;348:12011214.

11. Nagata H, Worobec AS, Oh CK, et al.Identication of a point mutation in thecatalytic domain of the protooncogene c-kitin peripheral blood mononuclear cells ofpatients who have mastocytosis with anassociated hematologic disorder. Proc NatlAcad Sci U SA 1995;92:1056010564.

12. Levine RL, Wadleigh M, Cools J, et al.Activating mutation in the tyrosine kinaseJAK2 in polycythemia vera, essentialthrombocythemia, and myeloid metapla-siawithmyelobrosis. Cancer Cell 2005;7:387397.

13. James C, Ugo V, Le Couedic JP, et al. Aunique clonal JAK2 mutation leading toconstitutive signaling causes polycythaemiavera. Nature 2005;434:11441148.

14. Kralovics R, Passamonti F, Buser AS, et al.A gain-of-function mutation of JAK2 inmyeloproliferative disorders. N Engl JMed2005;352:17791790.

15. Baxter EJ, Scott LM, Campbell PJ, et al.Acquired mutation of the tyrosine kinaseJAK2 in human myeloproliferative disor-ders. Lancet 2005;365:10541061.

TABLE 8. Innovative Therapies for Classic MPN

DRUG MAIN TARGETSIN CLINICALTRIAL

JAK2 selective inhibitors

INCB018424 JAK2 Yes

XL019 JAK2 Yes

TG101348 JAK2 Yes

Non-JAK2 selective inhibitors

CEP-701 (Leustartinib) FLT3 Yes

MK-0457 Aurora Kinase, FLT3, BCR-ABL Yes

Erlotinib EGFR Yes

ITF2357 Histone deacetylases Yes

Tipifarnib FT Yes

MPN indicates myeloproliferative neoplasm; FLT3, FMS-like tyrosine kinase 3;EGFR, epidermal growth factor receptor; FT, farnesyl transferase.



16. Scott LM, Tong W, Levine RL, et al. JAK2exon 12 mutations in polycythemia veraand idiopathic erythrocytosis.NEngl JMed2007;356:459468.

17. Pikman Y, Lee BH, Mercher T, et al.MPLW515L is a novel somatic activatingmutation in myelobrosis with myeloidmetaplasia. PLoSMed 2006;3:e270.

18. Pardanani AD, Levine RL, Lasho T, et al.MPL515 mutations in myeloproliferativeand other myeloid disorders: a study of1182 patients. Blood 2006;108:34723476.

19. Cazzola M, Skoda R. Gain of function, lossof controla molecular basis for chronicmyeloproliferative disorders. Haemato-logica 2005;90:871874.

20. Levine RL, Pardanani A, Tefferi A, Gilli-land DG. Role of JAK2 in the pathogenesisand therapy of myeloproliferative disor-ders. Nat Rev Cancer 2007;7:673683.

21. Delhommeau F, Pisani DF, James C,Casadevall N, Constantinescu S, Vainchen-ker W. Oncogenic mechanisms in myelo-proliferative disorders. Cell Mol Life Sci2006;63:29392953.

22. Tefferi A, Gilliland DG. Oncogenes inmyeloproliferative disorders. Cell Cycle2007;6:550566.

23. Saharinen P, Silvennoinen O. Thepseudokinase domain is required for sup-pression of basal activity of Jak2 and Jak3tyrosine kinases and for cytokine-induc-ible activation of signal transduction. J BiolChem 2002;277:4795447963.

24. FeenerEP,RosarioF,DunnSL, StanchevaZ,Myers MG, Jr. Tyrosine phosphorylation ofJak2 in the JH2 domain inhibits cytokinesignaling.MolCellBiol2004;24:49684978.

25. Mertens C, Darnell JE, Jr. SnapShot: JAK-STAT signaling. Cell 2007;131:612.

26. Starr R, Hilton DJ. Negative regulation ofthe JAK/STAT pathway. Bioessays 1999;21:4752.

27. Shuai K, Liu B. Regulation of JAK-STATsignaling in the immune system. Nat RevImmunol 2003;3:900911.

28. Sasaki A, YasukawaH, Shouda T, KitamuraT, Dikic I, Yoshimura A. CIS3/SOCS-3 sup-presses erythropoietin (EPO) signaling bybinding the EPO receptor and JAK2. J BiolChem2000;275:2933829347.

29. Stofega MR, Herrington J, Billestrup N,Carter-Su C. Mutation of the SHP-2 bind-ing site in growth hormone (GH) receptorprolongs GH-promoted tyrosyl phosphory-lation of GH receptor, JAK2, and STAT5B.Mol Endocrinol 2000;14:13381350.

30. Huang LJ, Constantinescu SN, Lodish HF.The N-terminal domain of Janus kinase 2is required for Golgi processing and cellsurface expression of erythropoietin recep-tor. Mol Cell 2001;8:13271338.

31. Royer Y, Staerk J, Costuleanu M, CourtoyPJ, Constantinescu SN. Janus kinasesaffect thrombopoietin receptor cell surfacelocalization and stability. J Biol Chem2005;280:2725127261.

32. Lu X, Levine R, Tong W, et al. Expressionof a homodimeric type I cytokine receptoris required for JAK2V617F-mediated trans-formation. Proc Natl Acad Sci U S A2005;102:1896218967.

33. Garcon L, Rivat C, James C, et al. Constitu-tive activation of STAT5 andBcl-xL overex-pression can induce endogenous erythroidcolony formation in human primary cells.Blood 2006;108:15511554.

34. Prchal JF, Axelrad AA. Letter: Bone-marrow responses in polycythemia vera.N Engl J Med 1974;290:1382.

35. Wernig G,Mercher T, Okabe R, Levine RL,Lee BH, Gilliland DG. Expression ofJak2V617F causes a polycythemia vera-like disease with associated myelobrosisin a murine bone marrow transplantmodel. Blood 2006;107:42744281.

36. Lacout C, Pisani DF, Tulliez M, GachelinFM, Vainchenker W, Villeval JL.JAK2V617F expression inmurine hemato-poietic cells leads to MPD mimickinghuman PV with secondary myelobrosis.Blood 2006;108:16521660.

37. Bumm TG, Elsea C, Corbin AS, et al.Characterization of murine JAK2V617F-positive myeloproliferative disease. Can-cer Res 2006;66:1115611165.

38. Zaleskas VM, Krause DS, Lazarides K, etal. Molecular pathogenesis and therapy ofpolycythemia induced in mice by JAK2V617F. PLoS ONE 2006;1:e18.

39. Tiedt R, Hao-Shen H, Sobas MA, et al. Ratioof mutant JAK2V617F to wild-type Jak2determines the MPD phenotypes in trans-genicmice. Blood 2008;111:39313940.

40. Vannucchi AM, Antonioli E, GuglielmelliP, Pardanani A, Tefferi A. Clinical corre-lates of JAK2V617F presence or alleleburden inmyeloproliferative neoplasms: acritical reappraisal. Leukemia 2008; 22:12991307.

41. Szpurka H, Tiu R, Murugesan G, et al.Refractory anemia with ringed sideroblastsassociated with marked thrombocytosis(RARS-T), anothermyeloproliferativecondi-tion characterized by JAK2 V617F muta-tion. Blood 2006;108:21732181.

42. Steensma DP, McClure RF, Karp JE, et al.JAK2 V617F is a rare nding in de novoacutemyeloid leukemia, but STAT3 activa-tion is common and remains unexplained.Leukemia 2006;20:971978.

43. Renneville A, Quesnel B, Charpentier A, etal. High occurrence of JAK2 V617 mutationin refractory anemia with ringed sidero-blasts associatedwithmarked thrombocyto-sis. Leukemia 2006;20:20672070.

44. Nishii K, Nanbu R, Lorenzo VF, et al.Expression of the JAK2 V617F mutation isnot found in de novo AML and MDS but isdetected in MDS-derived leukemia ofmegakaryoblastic nature. Leukemia 2007;21:13371338.

45. Vizmanos JL, Ormazabal C, Larrayoz MJ,Cross NC, Calasanz MJ. JAK2 V617Fmutation in classic chronicmyeloprolifera-tive diseases: a report on a series of 349patients. Leukemia 2006;20:534535.

46. Passamonti F, Rumi E, Pietra D, et al.Relation between JAK2 (V617F) mutationstatus, granulocyte activation, and consti-tutive mobilization of CD34 cells intoperipheral blood inmyeloproliferative dis-orders. Blood 2006;107:36763682.

47. Antonioli E, Guglielmelli P, Poli G, et al.Inuence of JAK2V617F allele burden onphenotype in essential thrombocythemia.Haematologica 2008;93:4148.

48. Pardanani A, Fridley BL, Lasho TL, Gilli-land DG, Tefferi A. Host genetic variationcontributes to phenotypic diversity inmyeloproliferative disorders. Blood 2008;111:27852789.

49. Skoda R. The genetic basis of myeloprolif-erative disorders. Hematology Am SocHematol Educ Program 2007;2007:110.

50. Nussenzveig RH, Swierczek SI, Jelinek J,et al. Polycythemia vera is not initiated byJAK2V617F mutation. Exp Hematol 2007;35:3238.

51. Theocharides A, Boissinot M, Girodon F,et al. Leukemic blasts in transformedJAK2V617F-positive myeloproliferativedisorders are frequently negative for theJAK2V617F mutation. Blood 2007;110:375379.

52. Campbell PJ, Baxter EJ, Beer PA, et al.Mutation of JAK2 in the myeloproliferativedisorders: timing, clonality studies, cytoge-netic associations, and role in leukemictransformation.Blood2006;108:35483555.

53. Pardanani A, Lasho TL, Finke C, HansonCA, Tefferi A. Prevalence and clinicopatho-logic correlates of JAK2 exon 12mutationsin JAK2V617F-negative polycythemiavera. Leukemia 2007;21:19601963.

54. Campbell PJ, Scott LM, Buck G, et al.Denition of subtypes of essential throm-bocythaemia and relation to polycythae-mia vera based on JAK2 V617F mutationstatus: a prospective study. Lancet 2005;366:19451953.

55. Pietra D, Li S, Brisci A, et al. Somaticmutations of JAK2 exon 12 in patients withJAK2 (V617F)-negative myeloproliferativedisorders. Blood 2007;111:16861689.

56. Pardanani A, Lasho T, Finke CM, et al.Extending JAK2V617F andMPL515muta-tion analysis to single myeloid coloniesand T and B lymphocytes. Stem Cells2007;25:23582362.

57. Chaligne R, James C, Tonetti C, et al.Evidence for MPL W515L/K mutations inhematopoietic stem cells in primitive my-elobrosis. Blood 2007;110:37353743.

58. Pardanani A, Lasho TL, Finke C,MarkovicSN, Tefferi A. Demonstration ofMPLW515K, but not JAK2V617F, in invitro expanded CD4 T lymphocytes.Leukemia 2007;21:22062207.

59. Kaushansky K. The molecular mecha-nisms that control thrombopoiesis. J ClinInvest 2005;115:33393347.

60. Guglielmelli P, Pancrazzi A, BergamaschiG, et al. Anaemia characterises patientswith myelobrosis harbouring Mpl muta-tion. Br J Haematol 2007;137:244247.

61. Beer PA, Campbell P, Scott LM, et al. MPLmutations inmyeloproliferative disorders:analysis of the PT-1 cohort. Blood 2008;112:141149.

62. Vannucchi AM, Antonioli E, GuglielmelliP, et al. Characteristics and clinical corre-lates of MPL 515WL/K mutation inessential thrombocythemia. Blood 2008;112:844847.

63. Williams DM, Kim AH, Rogers O, SpivakJL, Moliterno AR. Phenotypic variationsand new mutations in JAK2 V617F-negative polycythemia vera, erythrocyto-sis, and idiopathic myelobrosis. ExpHematol 2007;35:16411646.

64. Chaligne R, Tonetti C, Besancenot R, et al.Newmutations ofMPL in primitivemyelo-brosis: only the MPL W515 mutationspromote a G(1)/S-phase transition. Leuke-mia 2008;22:15571566.

65. Lasho TL, Pardanani A, McClure RF, et al.Concurrent MPL515 and JAK2V617F mu-tations in myelobrosis: chronology ofclonal emergence and changes in mutantallele burden over time. Br J Haematol2006;135:683687.



66. Fletcher S, Bain B. Diagnosis and treat-ment of hypereosinophilic syndromes.Curr Opin Hematol 2007;14:3742.

67. Buitenhuis M, Verhagen LP, Cools J,Coffer PJ.Molecularmechanisms underly-ing FIP1L1-PDGFRA-mediated myelopro-liferation. Cancer Res 2007;67:37593766.

68. Cools J, Stover EH, Boulton CL, et al.PKC412 overcomes resistance to imatinibin amurinemodel of FIP1L1-PDGFRalpha-induced myeloproliferative disease. Can-cer Cell 2003;3:459469.

69. Robyn J, Lemery S, McCoy JP, et al.Multilineage involvement of the fusiongene in patients with FIP1L1/PDGFRA-positive hypereosinophilic syndrome. Br JHaematol 2006;132:286292.

70. Golub TR, Barker GF, Lovett M, GillilandDG. Fusion of PDGF receptor beta to anovel ets-like gene, tel, in chronic my-elomonocytic leukemia with t(5;12) chro-mosomal translocation. Cell 1994;77:307316.

71. Wilkinson K, Velloso ER, Lopes LF, et al.Cloning of the t(1;5)(q23;q33) in a myelo-proliferative disorder associated with eo-sinophilia: involvement of PDGFRB andresponse to imatinib. Blood 2003;102:41874190.

72. Wang LC, Swat W, Fujiwara Y, et al. TheTEL/ETV6 gene is required specically forhematopoiesis in the bone marrow. GenesDev 1998;12:23922402.

73. Carroll M, Tomasson MH, Barker GF,Golub TR, Gilliland DG. The TEL/platelet-derived growth factor beta receptor (PDGFbeta R) fusion in chronic myelomonocyticleukemia is a transforming protein thatself-associates and activates PDGF beta Rkinase-dependent signaling pathways.Proc Natl Acad Sci U S A 1996;93:1484514850.

74. Valent P. Systemic mastocytosis. CancerTreat Res 2008;142:399419.

75. Cairoli R, Beghini A, Grillo G, et al. Prognos-tic impact of c-KITmutations in corebindingfactor leukemias: an Italian retrospectivestudy. Blood 2006;107:34633468.

76. Sihto H, Sarlomo-Rikala M, Tynninen O,et al. KIT and platelet-derived growthfactor receptor alpha tyrosine kinase genemutations and KIT amplications in hu-man solid tumors. J Clin Oncol 2005;23:4957.

77. Harir N, Boudot C, Friedbichler K, et al.Oncogenic Kit controls neoplastic mast cellgrowth through a Stat5/PI3-kinase signal-ing cascade. Blood 2008;112:24632473.

78. Vardiman JW, Harris NL, Brunning RD.The World Health Organization (WHO)classication of the myeloid neoplasms.Blood 2002;100:22922302.

79. Gilliland DG, Blanchard KL, Levy J, PerrinS, Bunn HF. Clonality in myeloprolifera-tive disorders: analysis by means of thepolymerase chain reaction. Proc Natl AcadSci U S A 1991;88:68486852.

80. Fialkow PJ, Gartler SM, Yoshida A. Clonalorigin of chronic myelocytic leukemia inman. Proc Natl Acad Sci U S A 1967;58:14681471.

81. Barr RD, Fialkow PJ. Clonal origin ofchronicmyelocytic leukemia.NEngl JMed1973;289:307309.

82. Adamson JW, Fialkow PJ, Murphy S,Prchal JF, Steinmann L. Polycythemiavera: stem-cell and probable clonal originof the disease. N Engl J Med 1976;295:913916.

83. Fialkow PJ, Jacobson RJ, Papayannopou-louT. Chronicmyelocytic leukemia: clonalorigin in a stem cell common to thegranulocyte, erythrocyte, platelet andmonocyte/macrophage. Am J Med 1977;63:125130.

84. Fialkow PJ, Denman AM, Jacobson RJ,LowenthalMN. Chronicmyelocytic leuke-mia. Origin of some lymphocytes fromleukemic stem cells. J Clin Invest 1978;62:815823.

85. Jacobson RJ, Salo A, Fialkow PJ. Agno-genic myeloid metaplasia: a clonal prolif-eration of hematopoietic stem cells withsecondary myelobrosis. Blood 1978;51:189194.

86. Martin PJ, Najfeld V, Hansen JA, PenfoldGK, Jacobson RJ, Fialkow PJ. Involve-ment of the B-lymphoid system in chronicmyelogenous leukaemia.Nature 1980;287:4950.

87. Fialkow PJ, Faguet GB, Jacobson RJ,Vaidya K, Murphy S. Evidence that essen-tial thrombocythemia is a clonal disorderwith origin in a multipotent stem cell.Blood 1981;58:916919.

88. Reeder TL, Bailey RJ, Dewald GW, TefferiA. Both B and T lymphocytes may beclonally involved in myelobrosis withmyeloid metaplasia. Blood 2003;101:19811983.

89. Delhommeau F, Dupont S, Tonetti C, et al.Evidence that the JAK2 G1849T (V617F)mutation occurs in a lymphomyeloid pro-genitor in polycythemia vera and idio-pathic myelobrosis. Blood 2007;109:7177.

90. Tefferi A, Gangat N, Wolanskyj AP, et al.20 yr without leukemic or brotic trans-formation in essential thrombocythemiaor polycythemia vera: predictors at diagno-sis. Eur J Haematol 2008;80:386390.

91. Tefferi A, Elliott MA, Pardanani A. Atypi-cal myeloproliferative disorders: diagno-sis and management. Mayo Clin Proc2006;81:553563.

92. Hehlmann R, Hochhaus A, Baccarani M.Chronic myeloid leukaemia. Lancet 2007;370:342350.

93. Schiffer CA. BCR-ABL tyrosine kinaseinhibitors for chronic myelogenous leuke-mia. N Engl J Med 2007;357:258265.

94. Kantarjian H, Schiffer C, Jones D, Cortes J.Monitoring the response and course ofchronic myeloid leukemia in the modernera of BCR-ABL tyrosine kinase inhibitors:practical advice on the use and interpreta-tion of monitoring methods. Blood 2008;111:17741780.

95. Elliott MA, Hanson CA, Dewald GW,Smoley SA, Lasho TL, Tefferi A. WHO-dened chronic neutrophilic leukemia: along-term analysis of 12 cases and acritical review of the literature. Leukemia2005;19:313317.

96. Elliott MA. Chronic neutrophilic leukemiaand chronic myelomonocytic leukemia:WHO dened. Best Pract Res Clin Haema-tol 2006;19:571593.

97. Gotlib J. Chronic eosinophilic leukemia/hypereosinophilic syndrome. Cancer TreatRes 2008;142:69106.

98. Tefferi A, Patnaik MM, Pardanani A.Eosinophilia: secondary, clonal and idio-pathic. Br J Haematol 2006;133:468492.

99. Simon HU, Plotz SG, Dummer R, Blaser K.Abnormal clones of T cells producinginterleukin-5 in idiopathic eosinophilia.N Engl J Med 1999;341:11121120.

100. Gotlib J, Cross NC, Gilliland DG. Eosino-philic disorders: molecular pathogenesis,new classication, and modern therapy.Best Pract Res Clin Haematol 2006;19:535569.

101. Kalac M, Quintas-Cardama A, Vrhovac R,Kantarjian H, Verstovsek S. A criticalappraisal of conventional and investiga-tional drug therapy in patients with hy-pereosinophilic syndrome and clonal eo-sinophilia. Cancer 2007;110:955964.

102. Garrett JK, Jameson SC, Thomson B, et al.Anti-interleukin-5 (mepolizumab) therapyfor hypereosinophilic syndromes. J Al-lergy Clin Immunol 2004;113:115119.

103. Sefcick A, Sowter D, DasGupta E, RussellNH, Byrne JL. Alemtuzumab therapy forrefractory idiopathic hypereosinophilic syn-drome. Br JHaematol 2004;124:558559.

104. Garcia-Montero AC, Jara-AcevedoM, Teo-dosio C, et al. KIT mutation in mast cellsand other bonemarrow hematopoietic celllineages in systemic mast cell disorders: aprospective study of the Spanish Networkon Mastocytosis (REMA) in a series of 113patients. Blood 2006;108:23662372.

105. Patnaik MM, Rindos M, Kouides PA, TefferiA, Pardanani A. Systemic mastocytosis: aconcise clinical and laboratory review. ArchPathol LabMed 2007;131:784791.

106. Pardanani A, Elliott M, Reeder T, et al.Imatinib for systemic mast-cell disease.Lancet 2003;362:535536.

107. Tefferi A, Verstovsek S, Pardanani A. Howwe diagnose and treat WHO-dened sys-temic mastocytosis in adults. Haemato-logica 2008;93:69.

108. Orazi A, GermingU. Themyelodysplastic/myeloproliferative neoplasms: myelopro-liferative diseases with dysplastic fea-tures. Leukemia 2008;22:13081319.

109. Hirsch-Ginsb

Documents

Wiley MPD 2009