nature reviews disease primers https://doi.org/10.

1038/s41572-025-00608-3

Primer Check for updates

Polycythaemia vera
Claire N. Harrison 1 , Tiziano Barbui 2, Prithviraj Bose 3, Jean-Jacques Kiladjian4, John Mascarenhas5,
Mary Frances McMullin6, Ruben Mesa7 & Alessandro M. Vannucchi 8
Abstract Sections

Polycythaemia vera (PV) is a haematological malignancy in the Introduction

myeloproliferative neoplasm family. PV is typically characterized Epidemiology

by erythrocytosis and often leukocytosis and thrombocytosis1. Mechanisms/pathophysiology
Clinical features include reduced life expectancy due to hazards of
Diagnosis, screening
thrombosis (often in atypical sites), haemorrhage and transformation to and prevention
myelofibrosis and less frequently to a form of acute myeloid leukaemia
Management
called blast phase. Almost two decades ago, the JAK2V617F mutation in
Quality of life
exon 14 of JAK2 was described, and is known to be present in more than
95% of patients with PV. Testing for the JAK2V617F mutation is used in the Outlook

diagnosis of PV, and the quantity of the mutation (that is, the variant
allele frequency) is linked to prognosis and the risk of complications.
As such, reduction of JAK2V617F variant allele frequency is currently being
evaluated as a treatment target. Recommendations for PV treatment
include control of vascular risk factors, therapeutic phlebotomy and
low-dose aspirin in all patients. Currently, patients at higher risk of
thrombosis (aged over 60 years and/or with a history of thrombosis)
are offered cytoreductive agents. Hydroxyurea or interferons remain
the preferred first-line cytoreductive agents, with the JAK1 and
JAK2 inhibitor, ruxolitinib, currently approved for the treatment of
patients who are resistant to, or intolerant of, hydroxyurea. Future
recommendations might be to treat the majority of patients with
these agents as long-term benefits of treatment begin to emerge.

1
Guy’s and St Thomas’ NHS Foundation Trust, London, UK. 2Research Foundation, Papa Giovanni XXIII Hospital,
Bergamo, Italy. 3The University of Texas MD Anderson Cancer Center, Houston, TX, USA. 4AP-HP, Hopital
Saint-Louis, Centre d’Investigations Cliniques CIC 1427, Université Paris Cité, Inserm, Paris, France. 5Tisch
Cancer Institute, Icahn School of Medicine at Mount Sinai, New York, NY, USA. 6Queen’s University, Belfast, UK.
7
Atrium Health Wake Forest Baptist Comprehensive Cancer Center, Winston-Salem, NC, USA. 8Center Research
and Innovation Myeloproliferative Neoplasms, University of Florence and AOU Careggi, Florence, Italy.
e-mail: [email protected]

Nature Reviews Disease Primers | (2025) 11:26 1

0123456789();:
Primer

Introduction with age-matched healthy individuals (Fig. 1). Both arterial and venous
The Philadelphia chromosome-negative myeloproliferative neo- thrombosis can occur in nearly 20% of patients either before or at the
plasms (MPNs), which include polycythaemia vera (PV), essential time of diagnosis, and the frequency of thrombotic events remains
thrombocythaemia, and primary myelofibrosis, are chronic neo- consistently higher in patients with PV than in a reference population
plastic blood disorders associated with excessive production of during follow-up, regardless of patient risk category6,7. Given the avail-
mature blood cells of the myeloid lineage1. PV is characterized by ability of new drugs such as ropeginterferon alfa-2b (a monopegylated
an abnormal increase in red blood cell (RBC) mass, often accom- interferon-α2b (IFNα2b)) and ruxolitinib (a JAK1 and JAK2 inhibitor), and
panied by leukocytosis and thrombocytosis. Clonal proliferation other drugs in development, hopefully enough favourable evidence will
of haematopoietic stem cells (HSCs) is driven by activating muta- accumulate in the near future to modify current treatment recommen-
tions in JAK2, which encodes Janus kinase 2, a non-receptor tyros- dations regarding disease management, with the possibility of positive
ine kinase that has important roles in cytokine and growth factor impacts on all of these disease manifestations8,9. The long-term nature
signalling. These activating JAK2 variants result in haematopoietic of clinical events in PV makes designing future clinical trials aimed
cytokine-independent signalling, thereby promoting clonal expan- at increasing progression-free survival challenging. One important
sion and myeloid differentiation2. Of note, the most frequent driver treatment goal is to have a favourable impact on overall survival, which
mutation in PV, JAK2V617F, is also commonly identified in a setting known is currently lower in patients with PV than the life expectancy in the
as clonal haematopoiesis of indeterminate potential (CHIP). CHIP is general age-matched and sex-matched population10,11.
characterized by the detection of somatic mutations in the peripheral This Primer covers the epidemiology, biology, therapy and future
blood of otherwise healthy people and occurs in at least 10% of adults research directions of PV, providing a comprehensive and updated
over the age of 70 years. It is well established that people with CHIP overview of this complex disease.
have increased rates of haematological malignancy, increased risk of
cardiovascular disease (CVD) and worse all-cause mortality compared Epidemiology
with those without CHIP3,4. Incidence and prevalence
The natural history of PV is marked by increased risk of thrombo- PV can occur in all age groups but is much more common in older
embolic events5, and in a small subset of patients, the disease evolves populations12. According to the Surveillance, Epidemiology and End
into myelofibrosis (an entity characterized by splenomegaly and tis- Results program data from the USA, the median age at PV diagnosis
sue fibrosis in the bone marrow, causing bone marrow failure, with a is 65 years13. Case ascertainment in epidemiological studies has been
survival of ~5 years), while an even smaller proportion (<5%) develop refined by the demonstration in 2005 that most patients (~95%) have
an acute myeloid leukaemia called blast phase. Myelofibrosis and blast the acquired (somatic) mutation JAK2V617F, indicating clonal disease14.
phase typically occur, on average, two or more decades after the diag- As this mutation can be detected in peripheral blood, case confirmation
nosis of PV. Patients with PV have reduced overall survival compared is greatly facilitated in epidemiological studies.

The MPN family of disorders

Severity of disease

Essential thrombocythaemia PV Primary or secondary myelofibrosis Post-MPN-AML

PV blood Healthy
blood

Increased red
blood cells

Natural history of PV
• Incidence: ~0.84 cases per 100,000
• Main problems:
• Thrombosis and/or haemorrhage
• Impact on quality of life
• Progression (15-year risk):
• Myelofibrosis: 6%
• Acute leukaemia: 5.5%
• Median overall survival of 20 years

Fig. 1 | Natural history of PV. Polycythaemia vera (PV) is a member of a group of post-PV myelofibrosis (this pathway of disease progression is more common
of disorders known as myeloproliferative neoplasms (MPNs); these include for PV than for essential thrombocythaemia). Myelofibrosis image courtesy of
essential thrombocythaemia, PV and myelofibrosis. A number of patients can Anna Green.
progress directly from PV to acute myeloid leukaemia (AML) or after a diagnosis

Nature Reviews Disease Primers | (2025) 11:26 2

0123456789();:
Primer

The incidence varies greatly between studies, many of which were

conducted in resource-rich countries, which could have influenced Box 1 | Management of pregnancy in PV
case identification. A systematic review and meta-analysis in 2015
showed reported annual incidence rates for PV ranging from 0.01 to Polycythaemia vera (PV) in pregnancy is high risk due to an
2.61 cases per 100,000 people. In this systematic review, the com- increased likelihood of thrombosis, miscarriage and other
bined annual incidence rate for PV was estimated to be 0.84 cases per complications. Management focuses on reducing thrombotic risk
100,000 people. This analysis included studies predominantly from for the mother, while minimizing harm to the fetus189.
Europe, but also from North America, Asia and Australasia15. An analysis • Low-risk pregnancies (no history of thrombosis, low
of registries in the European Union found that the annual incidence haematocrit): managed with low-dose aspirin and close
of PV varies from 0.4 to 2.8 cases per 100,000 people16. Sweden has monitoring.
comprehensive population-based registries with compulsory report- • High-risk pregnancies (history of thrombosis, prior pregnancy
ing, which show an age-standardized incidence of PV of 1.48 cases per complications): require prophylactic low-molecular-weight
100,000 person-years17. heparin (LMWH) in addition to aspirin.
As a chronic disease with prolonged survival, the prevalence of PV • Haematocrit control: target <38% (mid-gestational range), using
is likely to be much higher than the incidence, but in this regard there is phlebotomy if necessary. Cytoreductive therapy (for example,
little good quality information available. In the earlier noted systematic interferon-α) might be considered for high-risk patients.
review, five studies found that the prevalence of PV ranged from 0.49 Hydroxyurea is contraindicated in pregnancy.
to 46.88 cases per 100,000 people, so large variation is observed15. • Monitoring: regular fetal ultrasonography, Doppler studies and
A claims-based analysis in the USA showed a PV prevalence of 22 cases maternal blood counts.
per 100,000 (ref. 18). • Delivery planning: vaginal delivery is preferred unless obstetric
indications require caesarean section. Postpartum LMWH is
Mortality and morbidity recommended for at least 6 weeks to reduce thrombotic risk.
The natural history of PV is particularly marked by an increased risk of
thrombosis, disease transformation (to myelofibrosis or blast phase) Close collaboration between haematologists and obstetricians is
and early death19. Before any planned management for PV was cus- essential to optimize outcomes.
tomary, a retrospective study showed that 50% of untreated patients
with PV died within 18 months, mainly due to thromboembolic events.
The life expectancy of patients with PV is still currently lower than the long time-course of the disease. Observed clusters of MPNs have
the life expectancy observed in the age-matched and sex-matched prompted investigation for associated occupational exposures. For
general population11. The annual incidence of total major thrombosis example, in rural Pennsylvania, USA, a cluster was observed in popula-
is age-dependent: in patients aged 60, 70 and 80 years the incidence tions with occupational exposure to polycyclic aromatic hydrocarbons,
is 1.9%, 3.6% and 6.8% per year, respectively20. This finding is in line although in this cluster there was no clear indication of an association
with a contemporary population in which the annual thrombosis rate with cumulative exposure to these compounds27.
after diagnosis in patients with PV was approximately 2.6% per year21. Cohort studies in populations and case–control studies are the
The risk of transformation to myelofibrosis has been estimated at main implements to investigate any associations with MPN. A system-
between 15% and 25% over a patient’s lifetime, depending on individual atic search of the literature in 2011 highlighted known associations with
risk factors such as older age, higher JAK2 mutation allele burden and MPNs. In particular, PV was associated with family history of MPN and
additional somatic mutations, such as in ASXL1 or SRSF2. These muta- Jewish ethnicity. In cohort studies, some occupations were associated
tions are associated with aggressive disease phenotypes and predict with higher mortality from PV, including poultry workers, commercial
an increased likelihood of transformation22,23. PV can also transform pressmen and petroleum refinery workers, and in case–control stud-
into acute myeloid leukaemia (blast phase), or other leukaemias in ies, occupations such as rural sector jobs were associated with PV28.
2–14% of patients, with higher rates linked to mutational profiles (for Autoimmune conditions were associated with MPN; in particular,
example, oncogenic variants in TP53) and prior cytotoxic therapy giant cell arteritis was linked with PV28. Intriguingly, the incidence of PV
exposure, including agents such as busulfan or alkylating agents, in active blood donors in both Stockholm and France was higher than
radioactive phosphorus and potentially hydroxycarbamide. Blast expected29,30, but this association was not observed in later studies31.
phase arising from PV is associated with a poor prognosis, with survival The exploratory Myeloproliferative Neoplasms: An In-depth
often measured in months24. Case Control (MOSAICC) study identified several risk factors for MPN,
The natural history of PV is generally based on studies conducted including increasing household density, low childhood socioeconomic
in the older adult population, but small proportions of the teenage status and high number of pack-years smoking. Current smoking was
and young adult populations are also affected. In the past decade only associated with JAK2-positive PV and alcohol intake was inversely
these populations have been well-described in several articles; some associated with risk of PV32,33. A further, larger and definitive MOSAICC
important points include the potentially increased risks of splanchnic study is in progress and will report in the next few years, potentially
vein thrombosis in younger patients and the necessary consideration providing further insights. In a cohort study in the general population in
of special situations, for example pregnancy25,26 (Box 1). Denmark, daily and occasional or ex-smokers had a higher risk of having
an MPN diagnosis, including PV, compared with never-smokers34. The
Causes and risk factors association of MPN with smoking is a recurrent theme, and suggests
Environmental, lifestyle, medical and other factors are associated a role for inflammation in the development of MPNs. This hypothesis
with development of PV. The search for causes of MPN in general and is reviewed in a 2023 article; of interest, inflammation has also been
specifically for PV is difficult, owing to the rarity of the disorders and linked to the pathogenesis of thrombosis in MPN35. Obesity has been

Nature Reviews Disease Primers | (2025) 11:26 3

0123456789();:
Primer

associated with various blood cancers. In the Million Women Study the cytosolic portion of type I and type II cytokine receptors, which are
(a large cohort that included individuals with MPN), an increased risk devoid of intrinsic catalytic activity. In particular, JAK2 associates with
of MPN or myelodysplasia was associated with each 10 kg/m2 increase in the cytoplasmic domain of thrombopoietin receptor (MPL), erythro-
BMI (RR 1.37 for myeloid malignancy), but no information was obtained poietin receptor (EPOR), granulocyte colony-stimulating factor recep-
on specific MPN subtypes36. Another large study, the NIH-AARP Diet tor, growth hormone receptor and the IFNγ receptor 2. Upon cytokine
and Health Study, included a group of patients with PV, and identified binding to its cognate receptor, induction or reorientation of receptor
an increased risk of PV in those with high intake of sugar37. dimerization occurs, resulting in transphosphorylation-mediated acti-
Thus, multiple risk factors including environmental exposure, vation of the JAKs themselves, which in turn phosphorylate key tyros-
occupational exposure, medical conditions, smoking and inflamma- ine residues of the receptor cytoplasmic tail. These phosphorylated
tion, have been associated with the risk of developing and dying of PV, docks on the receptor and the JAKs, become scaffolds for signalling
but much further work remains to be done to elucidate the cause of PV, molecules, particularly members of the signal transducer and activa-
particularly at the individual level. Currently, the advice for patients tor of transcription (STAT) family, which once transphosphorylated by
would be that some genetic (ancestral) and lifestyle factors are likely to JAKs, dimerize and translocate to the nucleus to regulate gene expres-
be implicated and lifestyle factors in particular could be ameliorated, sion. JAK activation also initiates other signalling pathways such as the
for example, through reducing obesity and smoking. MAPK, PI3K and mTOR pathways (Fig. 2). All these proteins, once acti-
vated by phosphorylation, intervene at different levels in the control
Mechanisms/pathophysiology of cell survival, proliferation and differentiation43. Cytokine-mediated
Driver mutations in JAK2 activation of JAK2 is non-redundantly and critically required for nor-
The most prevalent mutation in PV is the somatic JAK2 gain-of-function mal haematopoiesis to proceed, as suggested by catastrophic effects
mutation, valine to phenylalanine at codon 617, located in exon 14, that (that is, embryonic lethality) of experimentally induced Jak2 knockout
is found in ~95% of patients2,38–40. Other heterogeneous insertions or in mice. Abnormal JAK–STAT signalling has been implicated across a
deletions (that is, ‘indels’) occur at JAK2 exon 12, that accounts for PV in number of myeloid and lymphoid malignancies and solid tumours44.
2–4% of patients41. The same JAK2V617F mutation is found in 50–60% of Activation of the JAK–STAT pathway could also promote reshaping
patients with essential thrombocythaemia and primary myelofibrosis, of the chromatin architecture and the profile of transcription factor
while exon 12 mutations are restricted to patients with PV alone. JAK2 binding through chromatin accessibility in normal cells44. The JAK2V617F
is one of the four genes in human genome encoding for JAKs: JAK1, mutation has been shown to directly impair, through phosphorylation,
JAK2, JAK3 and TYK2. protein arginine methyltransferase 5-mediated histone methylation45.
The JAKs have two kinase domains, of which one is catalytically Furthermore, JAK2-V617F mediates phosphorylation of tyrosine resi-
active at the protein carboxy terminus and the other is a pseudokinase due 41 on histone H3 (ref. 46), resulting in altered gene expression and
domain located upstream of the kinase domain that binds ATP, but contributing to enhanced myeloproliferation.
has no or a very weak ability to directly phosphorylate substrates42. The expression of JAK2V617F allele in different mouse models
The JAKs have a central role in many physiological processes, includ- (reviewed elsewhere47), which include mice transplanted with ret-
ing haematopoiesis, that involve cell-to-cell signalling through rovirally transduced bone marrow-derived haematopoietic pro-
cytokines and the downstream intracellular signalling cascade. JAKs are genitors, transgenic Jak2V617F mice and knock-in Jak2V617F mice,
non-covalently appended through an amino-terminal FERM domain to quite faithfully reproduced a PV phenotype with erythrocytosis,

JAK2 mutations in PV
EPO EPO e.g. JAK2V617F result in
constitutive activation

PI3K P RAS
JAK1 JAK2 HSPCs Megakaryocytes
P P
RAF P
P P
AKT P STAT STAT
MEK1/2 P
P STAT STAT P

mTOR
ERK1/2 P

MDM2
p53 BET
P PRMTS Myeloid cells Stromal cells
LSD1 P STAT STAT P
P • Inflammation
JAK2 H3 HP1α • Cytokines
• Disrupted iron homeostasis

Fig. 2 | Pathophysiology of polycythaemia vera. The complex interplay inflammation and iron homeostasis, all of which are known to be involved in
between haematopoietic stem and progenitor cells (HSPCs) in their niche, JAK2 the pathophysiology of polycythaemia vera. Arrows between cell populations
mutations and constitutive activation of downstream signalling pathways, indicate cellular crosstalk.

Nature Reviews Disease Primers | (2025) 11:26 4

0123456789();:
Primer

splenomegaly and variably increased leukocytes and platelets, rein- thrombosis history23. As discussed above, ASXL1 mutations have also
forcing the disease-initiating properties of the mutation. In this been associated with adverse outcomes. Interestingly, one hypothesis
regard, limiting dilution and single-cell transplantation in mice strik- is that, in some patients at least, the PV clone might originate from a
ingly demonstrated that single haematopoietic progenitors carrying single HCS before acquiring the JAK2V617F mutation. This mechanism was
the JAK2V617F mutation reconstituted animals and produced eryth- suggested from data showing that acute myeloid leukaemia transfor-
rocytosis and/or thrombocytosis48. Furthermore, compared with a mations in PV are frequently JAK2V617F mutation-negative69,70, yet share
JAK2V617F-mutant model, a transgenic exon 12 DelN542-E543 mouse the JAK2 haplotype. This finding underscores the effect of additional
model showed more dominant erythrocytosis, similar to human JAK2 mutations, and could have important future therapeutic impacts when
exon 12-mutated PV49–51. considering, for example, mutation-targeted therapies69. Of note,
The HSCs of patients with PV are clonal and the JAK2V617F driver karyotype abnormalities are found in about 12% of patients with PV and
mutation is found in long-term HSCs52. A conditional Jak2V617F knock-in include trisomy of chromosomes 8 or 9, alterations in chromosomes
mouse model demonstrated that initiation and long-term maintenance 5 and 7, and loss of the Y chromosome, which have all been associated
of PV, including myelofibrosis transformation over time, requires with reduced survival71–73.
expression of the mutant allele in the most immature, long-term HSC
population. By contrast, the PV phenotype was attenuated when Systemic effects
the Jak2V617F mutation was expressed in EPOR-expressing erythroid In PV, clonal dominance with the competitive advantage of the
precursors53. Another important observation is the requirement for malignant myeloproliferative clone, progression to myelofibrosis
MPL expression by HSCs for the induction of the PV phenotype by and systemic manifestations, are all different aspects of a profound,
JAK2V617F, beyond the absolute requirement of EPOR for the develop- chronic condition of hyperinflammation that is generated by the clone
ment of PV54. MPL is expressed on immature HSCs and is required for itself35 (Fig. 2). Sustained inflammation also contributes to the increased
their renewal, in addition to megakaryocyte differentiation and platelet vascular risk and atherosclerosis of patients with PV74. Circulating lev-
formation; impaired expression and trafficking of MPL in a preclinical els of acute phase inflammatory proteins (including high sensitivity
model of PV was demonstrated several years ago in seminal studies55. C-reactive protein (CRP) and pentraxin-3) are chronically elevated in
Germline variants in JAK2 and other genes (such as TERT, MECOM PV and these levels are correlated with increased JAK2V617F variant allele
and SH2B3 (ref. 56)) that predispose with low penetrance to sporadic frequency and increased risk of thrombosis75. Multiple cytokines were
and familial MPN have also been described (reviewed elsewhere57). In found to be elevated in the serum of patients with PV and associated
particular, a germline JAK2 46/1 (also known as GGCC) haplotype that with phenotypic traits and shortened survival76; a distinctive profile in
is common in the general population predisposes to the development relation to that of myelofibrosis was also highlighted76,77.
of PV with a nearly fourfold increased risk58–60, and was also reported Subtle differences in inflammatory signalling driven by the
to be associated with splanchnic vein thrombosis61. myeloproliferative clone might also exist, with IFNγ and STAT1-
directed gene expression being less prevalent in PV than in essen-
Additional pathogenic mutations tial thrombocythaemia 78. Inflammation might result from a
Beyond driver mutations in JAK2, patients with PV often harbour muta- combination of cell-autonomous-induced (via JAK2) and non-cell-
tions in epigenetic regulators (TET2, DNMT3A, ASXL1, IDH1 and IDH2) autonomous-induced (including cytokines such as IL-1α, IL-1β and TNF,
and splicing regulators (SF3B1, SRSF2, U2AF1 and ZRSR2), other signal- and non-malignant stromal cells) activation of the NFκB cascade79.
ling proteins (CBL), cell cycle regulators, transcription factors (TP53 Moreover, ablation of NFκB signalling in in vivo mouse models of PV
and NFE2), and others. These additional mutations occur in up to 53% alleviated the myeloproliferative phenotype. NFκB could therefore
of patients with PV62, particularly in TET2, ASXL1, SRSF2 and IDH2, which represent a target for therapy (for example, with BET inhibitors and
are also often found in age-dependent CHIP3. IRAK1–4 inhibitors)78. In particular, IL-1β-mediated inflammation in
The order of acquisition of mutations might critically contribute mice favoured the early clonal expansion of JAK2-mutant HSCs79. Of
to disease phenotype; for example, in a mouse model, inducing a TET2 note, elevated levels of IL-1β in the serum of patients with MPN, and IL-1
mutation first promoted a phenotype resembling essential thrombo- receptors on their haematopoietic progenitors and HSCs, correlated
cythaemia and inducing a JAK2 mutation first induced a phenotype with increased JAK2 variant allele frequency80. A substantial part of
that was more PV-like63. Furthermore, the combination of Jak2V617F the therapeutic activity of ruxolitinib (a JAK1 and JAK2 inhibitor) is
expression in mouse HSCs with concomitant Tet2 loss promoted the considered to be mediated by inhibition of inflammatory cytokines81,82.
fitness of Jak2-mutated cells, by conferring a strong functional com- Finally, the role of inflammation in the pathogenesis and progression
petitive advantage over normal haematopoiesis, highlighting func- of MPN is also exemplified by the link between cigarette smoke and the
tional consequences of a co-mutation profile in PV64. Patients with expansion of JAK2 variant allele frequency and TET2 myeloproliferative
PV who have a DNMT3A mutation might be less responsive to IFNα clones in preclinical models83.
therapy, and this mutation also selectively promoted the expansion
of DNMT3A–JAK2V617F co-mutated subclones in patients65, although this Dysregulated iron metabolism
finding is controversial66. Interestingly, such a resistance to IFN was Another level of deranged regulation in the pathobiology of PV is repre-
overcome by combination therapy of IFNα with azacytidine in mice sented by iron metabolism, that configures a condition of non-anaemic
bearing both mutations in haematopoietic cells. iron deficiency. Patients with PV at diagnosis are typically microcytic
Combinatorial effects of additional mutations with JAK2V617F could (RBCs smaller than normal) and iron-deficient (low serum levels of
also be involved in the promotion of progression to myelofibrosis67,68. iron and the iron storage protein ferritin, and low saturation of the
Mutations in SRSF2, documented in 2% of patients with PV, are con- iron-transport protein, transferrin) due to increased utilization of iron
sidered adverse mutations owing to their negative effect on survival by proliferating erythrocytes. Iron deficiency is further exacerbated
in a multivariable analysis, together with older age, leukocytosis and by repeated therapeutic phlebotomies that are needed to maintain

Nature Reviews Disease Primers | (2025) 11:26 5

0123456789();:
Primer

the recommended haematocrit level (a measure of the proportion Unusual sites of thrombosis, such as splanchnic veins (Budd–Chiari
of RBCs in blood) below 45%. Depletion of tissue iron storage could syndrome, portal or mesenteric veins) or cerebral sinuses, should raise
contribute to systemic symptoms, in particular fatigue and cognitive the suspicion for an underlying prothrombotic state, among which
impairment84,85. Whilst data are available that link iron deficiency to MPN, and PV in particular, is frequently found94.
thrombocytosis, whether a direct link to thrombosis in itself exists in
PV is unclear86. PV diagnostic criteria
In PV, evidence indicates that erythropoiesis continues at a high Two classification systems currently coexist for MPNs, both updated
rate despite iron deficiency and is associated with inappropriate sup- in 2022: WHO classification95 and the International Consensus Clas-
pression of hepcidin, the master regulator of iron homeostasis. In sification (ICC)1; however, these systems are similar and consider the
normal conditions, the serum levels of hepcidin are physiologically same key parameters for PV diagnosis96 (Fig. 3). There are three major
upregulated by increased circulating iron levels and mediators of criteria in both classifications: evidence of an erythrocytosis, charac-
systemic inflammation (particularly IL-6 and IL-1β)87, and are downregu- teristic histopathological changes in the bone marrow biopsy, and
lated by bone marrow erythropoiesis through the peptide hormone identification of a known mutation in the JAK2 gene.
erythroferrone released by erythroblasts88. Hepcidin restrains iron
availability by negatively regulating the activity of ferroportin (an iron Evidence of erythrocytosis. Unregulated production of RBCs leading
exporter channel expressed on duodenal enterocytes, hepatocytes to an absolute erythrocytosis is the distinctive feature of PV, among all
and macrophages); by reducing the export of intracellular iron, func- MPNs. Once a diagnosis of PV is suspected, an elevated haemoglobin
tional iron-deficient restriction of erythropoiesis occurs87. In spite of concentration (above 16.5 g/dl in men, above 16.0 g/dl in women)
suppressed levels of hepcidin in PV, systemic iron deficiency does not and/or haematocrit (above 49% in men, above 48% in women) are
resolve, implying dysregulated iron homeostasis that might originate deemed sufficient to diagnose PV according to the WHO classification
from concurrent systemic inflammation, elevated circulating levels of and the ICC (although local factors, such as comorbidities and altitude,
erythroferrone and/or aberrant tissue hypoxia-induced signalling89. should be accounted for), if the other criteria are fulfilled. The main
Administration of exogenous hepcidin reversed erythrocyto- change in these criteria compared with the previous versions is that
sis in JAK2V617F mice, possibly by restricting erythropoiesis following RBC mass measurement was removed from the diagnostic criteria in
increased iron sequestration in macrophages90. Also, treatment of the WHO classification, while it was maintained in the ICC. This test
patients with PV with ruxolitinib led to normalization of iron-related has become less accessible but it remains the only direct evidence of
parameters and increased serum hepcidin levels, possibly as a conse- true erythrocytosis, defined by an increased RBC mass above 125%
quence of drug-induced suppression of erythropoiesis85. Subtle differ- of the expected value. RBC mass is particularly useful in situations
ences might exist between JAK2V617F and JAK2 exon 12-mutated PV, as in where there is a notable expansion of the plasma volume (resulting in a
mice expressing exon 12-mutated Jak2, hepatic hepcidin expression was haemodilution that will induce an artificial decrease in the haematocrit
significantly lower than in Jak2V617F-mutated mice, while erythroferrone value)97. For example, in one study, half of the patients with splanchnic
expression was increased, with increased erythroferrone expression vein thrombosis had an absolute erythrocytosis demonstrated by the
also confirmed in analysis of erythroblasts from human primary PV49. RBC mass, while they did not reach the above-mentioned thresholds
A large population genome-wide association study showed that a sin- of haemoglobin or haematocrit for the diagnosis of PV98. In addition to
gle nucleotide polymorphism in HFE, which encodes the homeostatic enabling the diagnosis of PV, RBC mass can also help to set a therapeutic
iron regulator and is known to cause haemochromatosis, is associ- target in these patients with a haematocrit often spuriously lower than
ated in its heterozygous form with PV91. HFE influences the expres- 45%. It is important to be mindful of the limitations of relying on hae-
sion of hepcidin levels, and in different mouse models, endogenous moglobin or haematocrit, especially in the presence of iron deficiency,
hepcidin upregulation resulted in alleviation of the erythroid pheno- which will mask elevations in these parameters.
type, which might potentially explain the mechanism. Iron availabil-
ity for erythropoiesis may also be a modifier of disease phenotype; Histopathological changes in bone marrow. The second major
iron-deficient JAK2-mutant mice show increased platelet production criterion consists of specific findings in the bone marrow biopsy:
at the expense of erythropoiesis through primary involvement of an age-adjusted hypercellularity with trilineage proliferation
common pre-megakaryocyte–erythrocyte progenitors92. (panmyelosis), including prominent erythroid and granulocytic pro-
liferation, and particularly an increase in pleomorphic, mature mega-
Diagnosis, screening and prevention karyocytes without atypia (the latter being suggestive of myelofibrosis)
Signs and symptoms (Fig. 3). Bone marrow biopsy findings can also help reclassify patients
A diagnosis of PV can be made in three main circumstances. First, the with PV who initially presented with apparently isolated thrombo-
diagnosis can be made incidentally based on a peripheral blood count cytosis and red cell parameters below the thresholds for defining
performed for another reason or during a systematic check-up. Second, overt PV99,100. Interestingly, the ICC and WHO classification tolerate
a blood count can be prescribed owing to signs and symptoms that are avoiding bone marrow biopsy for diagnostic purpose in patients with
suggestive of PV, such as palpable splenomegaly (present in about 30% very high RBC parameters (haemoglobin and haematocrit above
of patients at diagnosis), aquagenic pruritus, microvascular symptoms 18.5 g/dl and 55.5% in men, and above 16.5 g/dl and 49.5% in women)
(visual disturbances including blurring and scotoma, dizziness, head- and JAK2 mutation. However, bone marrow biopsy at the time of PV
aches and erythromelalgia) or unexplained fatigue. Although none of diagnosis can offer important prognostic information, as the iden-
them is pathognomonic of PV, aquagenic pruritus and erythromelalgia tification of low-grade fibrosis is correlated with an increased risk of
(a sudden and transient erythema, with warmth or burning of digits) are unfavourable outcomes and evolution to post-PV myelofibrosis101.
characteristic93. Third, in approximately a quarter of patients, the diag- Still, bone marrow biopsy analysis can sometimes be difficult and
nosis is made after a vascular event, either thrombotic or haemorrhagic. significant inter-pathologist and intra-pathologist variability has

Nature Reviews Disease Primers | (2025) 11:26 6

0123456789();:
Primer

a
WHO 5th Edition ICC 2022
Major criteria Major criteria
• Elevated haemoglobin (>16.5 g/dl men; >16.0 • Elevated haemoglobin (>16.5 g/dl men; >16.0
g/dl women) or elevated haemocrit (>49% g/dl women) or elevated haemocrit (>49%
men; >48% women) men; >48% women) or increased red blood
• Bone marrow biopsy: hypercellularity with cell mass (>25% above mean normal
trilineage growth (panmyelosis), including predicted value)
erythroid, granulocytic and megakaryocytic • Bone marrow biopsy: age-adjusted
proliferation with pleomorphic, mature hypercellularity with panmyelosis and increase
megakaryocytes in pleomorphic mature megakaryocytes
• JAK2V617F or JAK2 exon 12 mutation without atypia.
• JAK2V617F or JAK2 exon 12 mutation
Minor criteria
• Subnormal erythropoietin level Minor criteria
• Subnormal erythropoietin level

b Hypercellular bone marrow with panmyelosis c Grade 1 fibrosis d Pleomorphic megakaryocytes

Reticulin CD61

Fig. 3 | Diagnostic criteria for polycythaemia vera. The diagnostic criteria and ICC is the presence of subnormal circulating levels of erythropoietin. Three
for polycythaemia vera (panel a) according to the WHO classification and the panels illustrating classic histological features are also included: hypercellularity
International Consensus Classification (ICC), both from 2022, include major with panmyelosis (panel b); increased reticulin and grade 1 fibrosis (in some
criteria relating to the erythrocytosis, bone marrow morphology and the patients) (panel c); and pleomorphic megakaryocytes highlighted with CD61
presence of the JAK2 mutation. A minor criterion for both WHO classification staining (panel d). Panels b–d, images courtesy of Anna Green.

been described102,103. Innovative approaches of digital pathology using (that should be less than 1%) is necessary and sufficient for the diagnosis
machine learning will undoubtedly provide considerable help for har- of PV, and quantification of the mutant variant allele frequency is not
monizing and standardizing several aspects of bone marrow biopsy required, although it is recommended, in this context by the ICC. Of
interpretation, such as megakaryocyte morphology and clustering, note, JAK2 mutations can be detected in individuals without evidence
or quantification of fibrosis104. of MPN or PV, referred to as CHIP; yet, this finding is associated with a
risk of vascular events and later development of MPN3,109,110. However,
Testing for pathogenic JAK2 mutation. The last major diagnostic cri- a measurement of variant allele frequency might be useful at diag-
terion for PV diagnosis is the presence of a mutation in JAK2; either the nosis for patient follow-up, as it can provide important prognostic
most frequent in exon 14 that leads to the JAK2-V617F-mutant protein information. For example, a high variant allele frequency above 50%
that is found in ~95% of patients with PV, or an indel in exon 12, found in has been correlated with an increased risk of disease evolution to
less than 5% of patients2,38–40,105. Genetic testing is usually performed on myelofibrosis111,112. On the other hand, therapies that can reduce the
whole peripheral blood. Of note, some subtle phenotypic differences JAK2V617F variant allele frequency are now available, including various
exist between patients with JAK2 mutations in exon 12 and those with formulations of IFNα113,114 and ruxolitinib115,116. Thus, an assessment of
JAK2V617F, including younger age at presentation and more marked the evolution of variant allele frequency during such therapies might
erythrocytosis in patients with exon 12 mutations106. Although not have relevance, as patients with PV who achieve a durable molecular
formally specified in the diagnostic criteria, in the rare patient with response have been demonstrated to have significant improvements
PV without exon 12 or 14 mutations, finding a non-canonical variant in their event-free, progression-free and overall survival116,117.
in JAK2 or SH2B3 (also known as LNK)107 can provide strong alternative In addition to variant allele frequency quantification, detection of
evidence for a PV diagnosis, but these patients are extremely rare108. additional non-driver mutations by next-generation sequencing (NGS)
Current diagnostic algorithms include only the presence of a is now frequently available118. NGS is not currently recommended in all
JAK2 mutation, regardless of the mutant allele burden. Therefore, a patients with PV at diagnosis, as most patients harbour a JAK2 mutation
qualitative assay detecting the mutation with a reasonable sensitivity that is sufficient for the diagnosis. Yet, additional non-driver mutations

Nature Reviews Disease Primers | (2025) 11:26 7

0123456789();:
Primer

Fig. 4 | Treatment approach for patients with

Evaluate toxicity and
response polycythaemia vera. The polycythaemia vera
Both haematological treatment paradigm begins with an accurate diagnosis
Aspirin and Choice of and molecular
and initiation of aspirin and phlebotomy to reduce
phlebotomy to cytoreductive
target haematocrit therapy haematocrit, and the risk stratification for both risk
of thrombosis and impact on survival. Patients at
Manage special high risk and selected patients at low risk receive
Establish Risk stratify Personalize therapy
situations
For thrombosis Manage symptoms cytoreductive therapy. Therapy also includes the
diagnosis E.g. pregnancy,
and survival and vascular risk measurement and optimization of symptoms and
surgery
vascular risk. Treatments are also personalized for
specific situations; over time treatment response and
toxicity as well as presence of disease progression
Consider disease must also be assessed.
progression

could clearly have prognostic relevance for the risk of disease progres- typically are eligible for cytoreductive therapy, in addition to thera-
sion to myelofibrosis or blast phase, as well as for improved assessment peutic phlebotomy and low-dose aspirin. Clinical practice guidelines
of the thrombotic risk119. In addition, NGS profiling could provide in do acknowledge special situations in which cytoreductive therapy
the future useful information for treatment decision; for example, might be indicated in low-risk patients, such as a frequent need for
treatment with JAK inhibitors could favour the selection of clones or poor tolerance of therapeutic phlebotomy, new thrombosis or
harbouring mutations in the RAS pathway. In addition, mutations in disease-related major bleeding, progressive thrombocytosis and/or
DNMT3A or TET2 might impair the response to ropeginterferon alfa-2b leukocytosis, splenomegaly or disease-related symptoms125,126. Man-
therapy65,68,120–122. agement of modifiable cardiovascular risk factors is often overlooked
but is of paramount importance.
Overall diagnostic strategy The definitions of low-risk and high-risk PV have traditionally
PV is suspected based on the presence of characteristic clinical fea- included below and above age 60–65 years and/or the presence of prior
tures or symptoms, or after a major thrombotic event (potentially in disease-related thrombosis or haemorrhage. Additional factors under
a rare territory, such as splanchnic veins or cerebral sinuses), along consideration are leukocytosis and JAK2 variant allele frequency. Stud-
with an increase in haemoglobin or haematocrit values as described ies have arrived at somewhat divergent conclusions as to the relative
above. In a patient with suspected PV, the first step should include contribution of leukocytosis to thrombosis risk. The European Col-
screening for JAK2 mutation and serum erythropoietin level, both laboration on Low-Dose Aspirin in Polycythemia Vera (ECLAP) study
tests being performed on a peripheral blood sample. In patients with (n = 1,638) identified a significant increase in the risk of thrombosis
clear erythrocytosis and a JAK2 mutation with a subnormal erythro- associated with a white blood cell (WBC) count >15 × 109/l, compared
poietin level, the diagnosis of PV can reliably be made. Importantly, with <10 × 109/l127. A time-dependent, multivariable subanalysis of
a subnormal erythropoietin level alone is insufficient for diagnosis. the Cytoreductive Therapy in Polycythemia Vera (CYTO-PV) study
Additional information can be collected in selected patients via bone examining WBC counts immediately preceding thrombotic events
marrow biopsy, quantitative evaluation of JAK2V617F variant allele fre- and organizing them into approximate quartiles found that the risk
quency and NGS for the screening for additional mutations. These of thrombosis steadily rose in patients with WBC counts >7 × 109/l,
additional tests that give prognostic information could potentially be becoming statistically significant when the WBC count was >11 × 109/l128.
proposed to patients diagnosed at a younger age, who will require a By contrast, however, a retrospective analysis that used group-based
very long follow-up and potentially several lines of treatment. Perform- trajectory modelling in 520 patients with PV followed at US academic
ing routine cytogenetic testing has been recommended at diagnosis, centres found that persistently elevated leukocyte trajectories were
although this testing is controversial as it does not currently affect not associated with thrombotic risk, although they were associated
treatment. with the risk of disease evolution129. In the past decade, the Prospec-
tive Observational Study of Patients with PV in US Clinical Practices,
Management a large, prospective, non-interventional study (n = 2,510) across the
The treatment of PV has traditionally focused on thrombosis risk US community and academic practices, showed that a WBC count
reduction in the absence of ‘disease-modifying’ therapies capable >11 × 109/l was significantly associated with the risk of thromboembo-
of affecting the underlying disease biology123 (Fig. 4). Patients are lism in both low-risk and high-risk patients with PV. The association
conventionally classified as low risk or high risk for thrombotic com- remained statistically significant for WBC counts >12 × 109/l in the
plications based on age and thrombosis history (other prognostic context of controlled haematocrit (≤45%)130. Elevated neutrophil-to-
models are included in Table 1). Patients younger than 60 years and lymphocyte ratios have been shown to predict venous thrombosis in
without previous thrombosis are considered at low risk of throm- PV and all cause mortality131,132.
bosis and standard management (as discussed in the next section) Taken together, these observations suggest that leukocytosis
includes low-dose aspirin and therapeutic phlebotomy to maintain the is probably a risk factor for thrombosis in PV, as well as CVD in the
haematocrit <45%, although this position might be changing towards non-PV population. Yet, prospective data indicating that normali-
more aggressive therapy as discussed below124. High-risk patients zation of leukocytes corrects this thrombotic risk are lacking. At
are those aged ≥60 years and/or with a history of thrombosis, and diagnosis, the ratio of JAK2V617F to the wild-type allele, expressed as

Nature Reviews Disease Primers | (2025) 11:26 8

0123456789();:
Primer

the JAK2 variant allele frequency, is usually higher in patients with PV 342 therapeutic phlebotomy only, demonstrated that with similar
than in those with essential thrombocythaemia or primary myelofi- follow-up (29.9 months for phlebotomy and 34.7 months for hydroxyu-
brosis, reaching 50% and more, and is associated with traits of more rea), the incidence of fatal or non-fatal cardiovascular events was 5.8
pronounced myeloproliferation and clinical manifestations111,133,134, versus 3 per 100 person-years, respectively (P = 0.002)138. A US Surveil-
including venous thrombosis135. The accumulation of mutated alleles lance, Epidemiology, and End Results–Medicare database study of
is one prevalent mechanism in the progression of PV to post-PV 820 older adults with PV found that 16.3% received neither phlebotomy
myelofibrosis112. nor hydroxyurea and this group had the worst outcomes in terms of
both thrombosis rate and survival, compared with those who received
Standard management therapeutic phlebotomy only, hydroxyurea only, or both139. Data are
Given that the most common complication associated with a diagnosis conflicting on whether the need for phlebotomy while receiving
of PV is a vascular event, it is logical that standard management should hydroxyurea affects thrombosis outcomes in PV140,141. In our practice,
include assessment and control of cardiovascular risk factors. The we routinely optimize the dosing of hydroxyurea or other cytoreduc-
seminal CYTO-PV study (n = 365) established a stringent haematocrit tive therapy to minimize or eliminate the requirement for phlebot-
goal of <45% in all patients with PV, regardless of the therapeutic modal- omy, in an effort to consistently maintain the haematocrit below 45%.
ity used136. In this study, despite an absolute difference of only 3.1% While an enduring controversy, hydroxyurea has never conclusively
between the median haematocrit values in the stringent haematocrit been shown to increase the risk of leukaemic transformation142–144.
(<45%) and less stringent haematocrit (45–50%) groups, a fourfold Exposure to hydroxyurea does, however, confer an increased risk of
reduction was observed in the incidence of the primary end point non-melanoma skin cancer145.
(major thrombosis or cardiovascular death) after a median follow-up
of 31 months in the stringent group. Currently, some experts prefer Interferons. IFNs have long been known to be associated with high
a lower haematocrit target of <42% in women, given the differences rates of haematological and molecular responses in patients with PV
in normal range between men and women, as well as a lower target (reviewed elsewhere146). Their specific mechanism of action is not fully
in patients with splanchnic vein thrombosis; however, these recom- elucidated but includes apoptosis, immune modulation and depletion
mendations are not evidence-based. Similarly, no current evidence of mutant HSC.
base exists for the appropriate target haematocrit in patients with PV In 2019, ropeginterferon alfa-2b, a monopegylated IFNα2b
who live at altitude. administered subcutaneously every 2 weeks, obtained regulatory
Of note, therapeutic phlebotomy to reduce haematocrit induces approval in Europe for the treatment of PV, marking the first licen-
iron deficiency with its attendant adverse effects and also dysregulates sure of an IFN formulation for the treatment of an MPN. US approval
iron metabolism. The frequency of phlebotomy is highly variable followed in 2021 (Table 2). The main appeal of IFNs lies in their
between patients but is usually most intense at the time of diagnosis. disease-modifying potential. Long-term follow-up of the phase III
Frequency of monitoring via blood count is therefore individually PROUD/CONTINUATION-PV trial that compared ropeginterferon
tailored. alfa-2b with hydroxyurea in patients with PV demonstrated a sig-
The ECLAP study established the efficacy and safety of daily nificant event-free survival advantage for ropeginterferon alfa-2b147.
low-dose aspirin in all patients with PV, unless contraindicated137. The label for ropeginterferon alfa-2b is both risk-agnostic and line of
therapy-agnostic, meaning for all patients deemed in need of therapy
Cytoreductive agents regardless of whether first-line or beyond. Results from the Low-PV
Hydroxyurea. Although hydroxyurea is the most widely used trial, which compared therapeutic phlebotomy and aspirin, with
cytoreductive agent in patients with PV, it has not been studied in phlebotomy and aspirin plus low fixed-dose ropeginterferon alfa-2b
a prospective, randomized controlled trial in a population with in patients with low-risk PV, showed superiority of the latter for the
contemporarily defined (for example, by detection of JAK2V617F) PV. composite end point of haematocrit control and absence of disease
However, a propensity score matching analysis of 1,042 patients progression, and could signal a paradigm shift in the management of
from the ECLAP study, of whom 681 received hydroxyurea and low-risk PV148,149.

Table 1 | Prognostic models for PV

Model Predicted outcome Key variables Risk stratification Ref.

Conventional risk model Thrombosis Age ≥60 years Low risk: neither factor 182
History of thrombosis High risk: one or both factors
Multiple factor-based prognostic Thrombosis Age ≥60 years Low risk: ≤1 point 183
score system for PV (MFPS-PV) Cardiovascular risk factors Intermediate risk: 1.5–2.5 points
High-risk mutations (e.g. DNMT3A, ASXL1, BCOR/BCORL1) High risk: ≥3 points
Previous thrombosis
Molecular international prognostic Overall survival Age >67 years Low risk: 0–1 points 23
scoring system for PV (MIPSS-PV) Leukocyte count ≥15 × 109/l Intermediate risk: 2–3 points
History of thrombosis High risk: ≥4 points
Presence of SRSF2 mutation
PV, polycythaemia vera.

Nature Reviews Disease Primers | (2025) 11:26 9

0123456789();:
Primer

Table 2 | Selected studies of pegylated interferons: efficacy and safety

Clinical trial Type Number Disease Treatments Clinical response Molecular Safety
(trial number) of state response
patients

Pegylated IFNα2aa

MPD-RC 111 (ref. 184) Prospective, 115 ET or PV Pegylated IFNα2a ET: CRb and PR at Median absolute AEs of any grade
(NCT01259817) international, 12 months in 43.1% change in were reported in
multicentre, and 26.2% of patients JAK2V617F VAF 90.4% of patients
open-label, phase II PV: CR and PR at was −6% (range and grade 3+ AEs
12 months in 22% and −84% to 47%) occurred in 43.8%
38% of patients in patients of patients
achieving CR
compared with
+4% (range
−18% to 56%)
in patients with
either a PR or no
response

MPN-RC 112 (ref. 150) Global, randomized, 168 ET or PV HU or pegylated ET: CRb at 36 months Median greatest 96% patients
(NCT012259856) phase III IFNα2a was 17% for HU and 40% change from were assessed for
for IFNα baseline in AEs. Grade 3 or
PV: CR at 36 months JAK2V617K VAF was higher occurred
was 17% for HU and 29% −5.3% for HU and in 37% patients
for IFNα −10.7% for IFNα (HU 28%, IFNα
46%)

Ph2 in ET or PV185 Prospective, 83 ET or PV Pegylated IFNα2a ET: CRc and PR in 73% Of the Severe
(NCT00452023) open-label, and 3% of patients JAK2V617F-positive haematological
single-centre, phase II PV: CRc and PR in 77% patients AEs (grade 3 or
and 7% of patients evaluable for 4) or recurrent
molecular events occurred
response (n = 55), in 89% of patients
the JAK2V617F VAF with PV and 83%
was reduced of patients with ET
in 63%

Pegylated IFNα2b

NMPD Ph2 in PV or ET Prospective, 42 ET or PV Pegylated IFNα2b At 2 years, 45% had At 2 years, 36% The majority of
open-label, phase II achieved complete of patients who AEs were WHO
platelet response were positive for grade 1 or 2; 42%
PRV-1 and had of patients who
samples available completed 2
achieved years of therapy
normalized PRV-1 reported any side
expression effect at the end
of the study

DALIAH186 Multicentre, 202 ET, PV, HU, pegylated At 3 years, A partial Treatment DC
(NCT01387763) randomized, pre-MF IFNα2a or clinicohaematological molecular for any reason at
controlled, phase III or PMF pegylated IFNα2b responsed was achieved response was 3 years was 21%
in 71% of HU-treated observed in 23% in HU-treated
patients, 43% of of HU-treated patients, 53%
IFNα-treated patients patients, 29% in IFNα-treated
(>60 years) and 41% of of IFNα-treated patients
IFNα-treated patients patients (>60 (>60 years)
(≤60 years) years) and in 28% and 56% in
of IFNα-treated IFNα-treated
patients patients
(≤60 years) (≤60 years)

Ropeginterferon alfa-2b

PEGINVERA187 Multicentre, 51 PV Ropeginterferon Best observed Best observed 94.1% of patients

(NCT01193699) open-label, alfa-2b haematological individual reported at
dose-escalation, response in efficacy set molecular least one AE,
phase I/II was CHR in 64.3% of response was the majority of
patients and PR in 33.3% CMR in 28.6% of which were mild
patients and PR in or moderate in
45.2% severity

Nature Reviews Disease Primers | (2025) 11:26 10

0123456789();:
Primer

Table 2 (continued) | Selected studies of pegylated interferons: efficacy and safety

Clinical trial Type Number Disease Treatments Clinical response Molecular Safety
(trial number) of state response
patients

Ropeginterferon alfa-2b (continued)

PROUD-PV Multicentre, 169 PV HU–BAT or At year 6, 81.4% and JAK2V617F VAF <1% Event-free
(NCT01949805) open-label, ropeginterferon 60.0% of patients in at 6 years was survivale over
CONTINUATION-PV113,155 randomized, alfa-2b the ropeginterferon achieved in 20.7% ≥6 years was
(NCT02218047) controlled, phase III alfa-2b and HU groups, patients in the significantly
respectively, did not ropeginterferon higher in the
require phlebotomies alfa-2b arm ropeginterferon
to maintain Hct <45% compared alfa-2b group
with 1.4% in than in the
the HU–BAT HU–BAT group
arm (patients (risk events
with baseline reported in 5/95
VAF >10%) vs 12/74 patients,
respectively)
Low-PV148,149,188 Multicentre, 127 PV Phlebotomy + After 2 years of Ropeginterferon Treatment-related
(NCT03003325) open-label, acetylsalicylic acid treatment, 82.7% of alfa-2b AEs were
randomized, phase II with or without patients who responded responders reported in 55%
ropeginterferon to treatment in the achieved a and 6% of the
alfa-2b ropeginterferon alfa-2b 23.1% decrease ropeginterferon
group had achieved in JAK2V617F at alfa-2b and
responsef compared 2 years compared standard
with 59.4% in the with a 15.4% treatment arms,
standard treatment arm increase in respectively
the standard
treatment arm
AE, adverse event; BAT, best available therapy; CHR, complete haematological response; CMR, complete metabolic response; CR, complete response; DC, discontinuation; ET, essential
thrombocythaemia; Hct, haematocrit; HU, hydroxyurea; IFN, interferon; PR, partial response; pre-MF, pre-myelofibrosis; PMF, primary myelofibrosis; PV, polycythaemia vera; VAF, variant allele
frequency. aThe DALIAH study also involved patients treated with pegylated IFNα2a. bComplete response was defined as correction of the platelet count to <400 × 109/l, haematocrit to <45%
without phlebotomy (for patients with PV only) and white blood cell count to <10 × 109/l; resolution of splenomegaly; and resolution of disease-related symptoms (defined as microvascular
disturbances, headache and pruritus). cComplete haematological response was defined as normalization of blood counts (in essential thrombocythaemia, platelets ≤440 × 109/l; in PV,
haemoglobin <15.0 g/l without phlebotomy) with complete resolution of palpable splenomegaly/symptoms in the absence of a thrombotic event. A partial response required at least a 50%
reduction in the platelet count for essential thrombocythaemia, or a 50% reduction in the rate of phlebotomies, or a 50% reduction in spleen size by palpation for PV. dClinicohaematological
response assessment was performed by central review according to the modified 2009 European LeukaemiaNet (ET, PV and pre-MF) and the 2005 European Myelofibrosis Network criteria
(PMF). eRisk events: disease progression, death and thromboembolic events. fResponse defined as: haematocrit <45% and no need for additional cytoreductive treatment. Reprinted from
ref. 146, CC BY NC 4.0.

Most previous studies of IFN formulations have used pegylated in patients with PV, achievement of sustained complete haemato-
IFNα (Table 2). The Myeloproliferative Disorders Research Consor- logical remission and reduction in the JAK2V617F variant allele frequency
tium 112 study in previously untreated patients with high-risk PV or to < 10% were predictive of being able to safely discontinue therapy
essential thrombocythaemia requiring cytoreduction did not show with pegylated IFNα2a153. While better tolerated than non-pegylated
meaningful differences between pegylated IFNα2a and hydroxyu- IFN, the long-term tolerability of pegylated IFNα2a remains problem-
rea after 12 months of treatment150. Although greater reductions in atic, with a 22% discontinuation rate found in one study154. However,
JAK2V617F variant allele frequency were observed with pegylated IFNα2a, ropeginterferon alfa-2b, has a less frequent dosing schedule and more
hydroxyurea led to more histopathological responses in bone mar- favourable pharmacokinetic profile than older IFN therapies; therefore
row biopsies, underscoring a possible disconnect between these ropeginterferon alfa-2b might have a superior tolerability profile155.
two outcome measures. However, this trial was terminated early and Consistent with consensus guidelines, we prefer the use of IFN prepa-
the results should be interpreted with this knowledge. Moreover, rations in younger patients with PV, given their disease-modifying
improved myelofibrosis-free and overall survival with pegylated potential and lack of teratogenicity and carcinogenicity. The ongoing
IFNα were found in a large single-institution retrospective study in ECLIPSE PV trial (NCT05481151) will hopefully clarify the optimal dosing
patients with PV, compared with hydroxyurea or phlebotomy only151. strategy for ropeginterferon alfa-2b.
Molecular analyses of data from the DALIAH trial in PV identified
treatment-emergent DNMT3A mutations as a possible mechanism of Ruxolitinib. The JAK1–2 inhibitor ruxolitinib is approved for the treat-
resistance to pegylated IFNα65; JAK2-mutated clones seem more sensi- ment of PV in the setting of hydroxyurea resistance or intolerance (see
tive to IFN therapy than TET2-mutated or CALR-mutated clones65,152. In Box 2 for the European LeukaemiaNet definition156). Patients with
patients with PV or essential thrombocythaemia, increasing genomic PV who develop resistance to hydroxyurea have worse survival and
complexity of the clone was shown to correlate with poorer responses higher rates of leukaemic transformation than those with no treatment
to IFN therapy122. resistance157. The approval of ruxolitinib was based on the results of
Given the disease-modifying potential of IFNs, there is consider- the RESPONSE trials in controlling haematocrit levels, reducing sple-
able interest in the possibility of treatment-free remission. In one study nomegaly and improving symptoms (Table 3) in patients with PV who

Nature Reviews Disease Primers | (2025) 11:26 11

0123456789();:
Primer

front-line setting with a primary end point of event-free survival. Of note,

Box 2 | European LeukaemiaNet definition 5-year follow-up of the RESPONSE studies showed excellent durability
of hydroxyurea resistance and intolerance of response to ruxolitinib and no new safety signals115,163. Weight gain,
in polycythaemia vera herpes zoster reactivation and non-melanoma skin cancers are impor-
tant adverse events associated with ruxolitinib. In our practice, we offer
Definitions of hydroxyurea resistance and intolerance in inactivated shingles vaccination to all patients treated with ruxolitinib.
polycythaemia vera are provided in refs. 126,156.
Novel agents in clinical development
Hydroxyurea resistance after 3 months of treatment Novel agents in clinical development for the treatment of PV include
• Need for phlebotomy to keep haematocrit <45%. hepcidin-modulating agents (rusfertide, sapablursen and divesiran),
• Uncontrolled myeloproliferation (that is, platelet count the histone deacetylase inhibitor, givinostat, and the lysine-specific
>400 × 109/l and white blood cell count >10 × 109/l). demethylase 1 inhibitor, bomedemstat. All of these agents are cur-
• Failure to reduce massive splenomegaly by more than 50% rently in active phase I–III clinical trials (Table 4). Hepcidin acts as
as measured by palpation, or failure to completely relieve the master regulator of iron availability, through inhibition of the
symptoms related to splenomegaly. Here, splenomegaly is iron transporter ferroportin expressed by enterocytes, hepatocytes
defined as the organ extending by more than 10 cm from the and macrophages164. Serum levels of hepcidin are inappropriately
costal margin. depressed in patients with PV, and in a JAK2V617F mouse model of PV,
administration of a mini-hepcidin mimetic resulted in normalization of
Hydroxyurea intolerance haematocrit and splenomegaly165. This concept was then evaluated in
• Absolute neutrophil count <1.0 × 109/l or platelet count <100 × 109/l a phase II trial of self-administered, weekly subcutaneous injection of
or haemoglobin <100 g/l at the lowest dose of hydroxyurea rusfertide (PTG-300) in a dose-finding stage followed by a double-blind,
required to achieve a complete or partial clinicohaematological randomized withdrawal period with placebo over 12 weeks, in
response. Complete response is defined as: haematocrit <45% 70 patients with either low-risk or high-risk PV, approximately half of
without phlebotomy, platelet count ≤400 × 109/l, white blood cell whom were receiving concurrent cytoreductive therapy166. A response,
count ≤10 × 109/l, and no disease-related symptoms. Partial which was defined as haematocrit control and the absence of phle-
response is defined as: haematocrit <45% without phlebotomy, botomy, was attained in 60% versus 17% in the placebo arm (P = 0.002).
or response in three or more of the other criteria. Improvements in symptoms and iron indices were observed with rusfer-
• Presence of leg ulcers or other unacceptable tide but no improvements in leukocyte or platelet counts were noted,
hydroxyurea-related non-haematological toxicities, such as and changes in JAK2V617F variant allele frequency were not reported. An
mucocutaneous manifestations, gastrointestinal symptoms, ongoing randomized phase III trial has completed accrual, and initial
pneumonitis or fever at any dose of hydroxyurea. results are anticipated in 2025 (NCT05210790).
A number of other therapeutic interventions converging on the
hepcidin pathway are also under early-phase clinical evaluation. The
were resistant or intolerant to hydroxyurea. Thromboembolic events antisense oligonucleotide, sapablursen, targets TMPRSS6 mRNA (that
were not captured as efficacy end points in these trials but rather only encodes a transmembrane serine protease matriptase-2), which nega-
as safety end points, and numerically favoured ruxolitinib treatment tively regulates the hepcidin coreceptor, hemojevulin. The net effect
over best available therapy158,159. Thus, prospective data confirming the of sapablursen is upregulation of hepcidin expression, and this subcu-
potential of ruxolitinib to reduce thromboembolic events in patients taneous injectable is being evaluated in an open-label phase IIa trial in
with PV are lacking, although a meta-analysis suggested ruxolitinib phlebotomy-dependent patients with PV (NCT05143957). Divesiran,
does indeed reduce thromboembolic events in PV160. an N-acetylgalactosamine conjugated 19-mer short interfering RNA of
In 2023, the investigator-initiated MAJIC-PV trial in patients with TMPRSS6, is also being evaluated in phlebotomy-dependent patients
PV and hydroxyurea resistance or intolerance comparing ruxolitinib with PV in a phase I–II trial (NCT05499013).
and the best available therapy found significant improvement in Givinostat is an oral pan-histone deacetylase inhibitor that modu-
event-free survival, and particularly thromboembolism-free survival, lates acetylation status of both histone and non-histone proteins (such
with ruxolitinib116. This trial also showed a higher rate of molecular as the heat shock protein 90 molecular chaperone); this drug reduces
response in the ruxolitinib-treated patients, which correlated with JAK–STAT signalling and selectively reduces JAK2-V617F protein levels
improved event-free survival. A mean reduction in JAK2V617F variant allele in MPN cell lines and primary cells167. Givinostat has been extensively
frequency from 52.2% to 40% after up to 4 years of therapy in patients studied in a number of early-phase trials in PV as a monotherapy and
with PV randomized to ruxolitinib was also found in the RESPONSE in combination with hydroxyurea, with response rates that range from
trial161. In 2023, the RUXOBEAT trial of ruxolitinib versus best avail- 50% to 80% in controlling blood counts and improving symptoms168–171.
able therapy in the frontline setting failed to meet its primary end Myelosuppression and gastrointestinal toxicity are well characterized
point of clinicohaematological complete response at month 6, but class effects, seen in approximately 50% of treated patients and can lead
the definition of complete response required a strict score of zero for to dose modifications and even discontinuation in some patients. The
each of four symptoms: headache, pruritus, dizziness and concen- GIV-IN global, randomized, phase III trial in high-risk patients with PV
tration problems162. Ruxolitinib, as demonstrated in the RESPONSE within 3 years of diagnosis will compare givinostat and hydroxyurea
and MAJIC-PV trials, is a particularly efficacious drug for the control with a primary end point of complete haematological remission and
of cytokine-mediated and splenomegaly-related symptoms, but the normal spleen size at week 48 (NCT06093672).
RUXOBEAT study probably set an unachievably high bar. The ongoing Finally, two novel JAK inhibitors targeting either the JAK2V617F
MITHRIDATE study (ISRCTN 12885480) is assessing ruxolitinib in the mutation or the pseudokinase domain are currently in development

Nature Reviews Disease Primers | (2025) 11:26 12

0123456789();:
Primer

(NCT06343805 and NCT06313593). The hope is that these drugs might medication-related toxicities, the financial burdens of therapy, the
have fewer off-target adverse effects, such as suppressed immunity or complexity and hassle of medical care, travel needed for medical care,
skin cancers, among others. Data are not yet available. anxiety over uncertain health, the effects of disease on activities of
The efficacy of therapies in clinical development for PV will need daily living or desired activities or travels, among others. Symptoms
to be proven in terms of classic phase III trial end points such as hae- related to a disease such as PV can clearly affect HR-QOL, sometimes
matological response and symptom benefit, and importantly must substantially and, if severe enough, can be the biggest factor (for exam-
ensure tolerability, as these agents are anticipated to be administered ple, severe fatigue, pruritus or night sweats) that might dramatically
over the long term. Additionally, cost to the patient and health-care affect an individual.
systems will need to be considered as we add to our therapeutic arma- The myelofibrosis symptom assessment form (MF-SAF) was devel-
mentarium. Drug development should be focused on translation of oped to quantify symptoms in patients with myelofibrosis173, based
mechanism-based anticlonal treatments with correlative evidence on survey data gathered from patients with myelofibrosis, and has
of disease modification (driver mutation molecular remission, bone evolved to MF-SAF 4.0, the current standard. Recognizing early on that
marrow histopathological response) and ultimately thrombosis-free symptoms are present in essential thrombocythaemia and PV that are
and progression-free survival. less relevant in myelofibrosis, a broader based MPN symptom assess-
ment form (MPN-SAF)174 was developed, and has been validated in
Quality of life dozens of languages and used in numerous phase III clinical trials. These
The constellation of challenges that patients with PV face includes patient-reported outcomes (PROs) forms have been highly validated
a range of symptoms, quality of life (QoL) and health issues (Fig. 5). to be: first, understandable by patients; second, able to reproducibly
Indeed, even in Sir William Osler’s first descriptions of the disorder, measure individual symptoms; and third, able to measure change in
there was recognition of plethora, reddening of the skin and nota- response to therapy. Over time, we have also seen great consistency
ble pruritus, amongst other challenges. The first efforts to improve in the prevalance and ranking of symptoms across countries and lan-
the quantification of these symptoms were driven by patient advo- guages, with some effects of culture on absolute symptom severity
cates with MPN who approached investigators. Comprehensive between groups (but with the ability to measure change preserved).
cross-sectional survey research of patients with MPN demonstrated Given that each patient acts as his or her own control (measuring the
fatigue as the most common symptom and aquagenic pruritus as a clas- change compared with baseline in individual symptoms or in aggre-
sic symptom172. Challenging symptoms can be a direct consequence of gate), a total symptom score was developed. Responses in trials have
erythrocytosis affecting vascular flow (for example, headaches, includ- been variable, ranging from the percentage of patients with a 50%
ing migraines with aura, or challenges with concentration), a conse- reduction in total symptom score compared with baseline (typically a
quence of splenomegaly (in a subset of patients) with early satiety, pain minimal evaluable baseline), or a measurement of change in the single
and abdominal fullness, or metabolically derived (for example, drench- most bothersome symptom(s).
ing night sweats). Other symptoms that can be noted in PV include The National Comprehensive Cancer Network guidelines recom-
weight loss, bone pain, fever and more advanced splenomegaly-related mend that PV symptoms be measured by the MPN-SAF at baseline and
symptoms, and in a patient with PV these might signal progression to serially with management175. The presence of problematic symptoms
myelofibrosis. is used in the guidelines as a threshold for initiation of cytoreduc-
Health-related QoL (HR-QOL) looks at QoL as affected by tive therapy in those patients on phlebotomy and aspirin alone, or
health, and can include symptoms from a disease, but also include for changing cytoreductive therapy to a second or subsequent line.

Table 3 | Pivotal phase III randomized controlled trials of ruxolitinib in polycythaemia vera

Trial Population Basic design and Major efficacy end Salient points about Long-term results
sample size points safety

RESPONSE115,158 HU-resistant or Ruxolitinib (starting At week 32, 21% vs Herpes zoster 88% crossed over to ruxolitinib
HU-intolerant with dose 10 mg twice daily) 1% for composite (grade 1 or 2) in 6.4% (by week 80); 5-year probability
splenomegaly (≥450 cm3), vs standard therapy primary outcome, in ruxolitinib group of maintaining composite primary
phlebotomy-dependent (1:1, n = 222), primary 60% vs 20% for Hct vs 0% in standard outcome 74%, of CHR 55%, of overall
end point Hct control control, 38% vs 1% for therapy group; clinicohaematological response 67%;
plus SVR35 at week 32; SVR35, 24% vs 9% for grade 3–4 infection anaemia most common AE in patients
crossover permitted CHR, 49% vs 5% for in 3.6% vs 2.7%; TE receiving ruxolitinib
after week 32 TSS50 events, 1 vs 6 through
week 32

RESPONSE-2 HU-resistant or Ruxolitinib (starting At week 28, Hct Anaemia and 77% crossed over to ruxolitinib (by
(refs. 159,163) HU-intolerant without dose 10 mg twice control achieved in thrombocytopenia week 80); at week 260, 22% of patients
palpable splenomegaly, daily) vs best available 62% vs 19%, CHR in the most frequent in the ruxolitinib group had achieved
phlebotomy-dependent therapy (1:1, n = 149), 23% vs 5%, TSS50 haematological AEs durable Hct control (estimated
primary end point Hct in 45% vs 23% (no grade 3–4 with median duration not reached); median
control at week 28; ruxolitinib) Hct in ruxolitinib group remained
crossover permitted <45% through 5 years of follow-up
after week 28
AE, adverse event; CHR, complete haematological response; Hct, haematocrit; HU, hydroxyurea, SVR35, spleen volume reduction by ≥35%; TE, thromboembolic; TSS50, total symptom score
reduction by ≥50%.

Nature Reviews Disease Primers | (2025) 11:26 13

0123456789();:
Primer

Of note, what is considered a problematic symptom is subjective and Evolving prognostication will better identify those patients with PV,
identified after a discussion between the treating physician and patient. possibly even all patients, who should probably be treated with the
Improvement in PV-related symptoms was essential for assessing newer, less toxic, potentially more potent, cytoreductive agents that
responses to hydroxyurea and pegylated IFN in PV176, for trials investi- are currently available.
gating ropeginterferon alfa-2b in low-risk PV148, and for demonstrating Challenges or indeed opportunities in the future include the
the benefit of ruxolitinib in improving PV symptoms116. Improvement need to understand how to prioritize therapy targets; that is, com-
in symptoms has also been demonstrated in trials of hepcidin mimet- plete haematological response or molecular response or both. Work
ics (such as rusfertide166). Baseline and serial assessment of PV-related is needed to harmonize our abilities to robustly measure changes in
symptoms is as important for management as are serial complete the JAK2V617F variant allele frequency. We also need to understand what
blood counts, and affects the use of standard approaches as well as factors affect attainment of molecular response and what happens to
clinical trials. non-driver mutations. For example, we should investigate whether
clinicians should aim to reduce the JAK2V617F variant allele frequency
Outlook within a year or whether it is sufficient to treat with the aim of achiev-
PV is an acquired myeloproliferative clonal disorder caused by a ing a target of, say, 50% reduction within 3 years. Of note, in defining
mutant JAK2 HSC clone. In this condition, the peripheral blood genetic molecular response, a 50% reduction is a much more modest threshold
mutation profile reflects that of the HSC and can be repeatedly sam- than that currently used in other myeloid malignancies, such as chronic
pled; thus, we have made considerable progress in our understanding myeloid leukaemia. If the 50% reduction is achieved, then we could ask
of the pathogenesis and progress of this condition. The description of whether the aim should be for stabilization or further reduction. A bet-
JAK2-mutated CHIP3, and also elegant work showing the likely occur- ter definition of true minimal residual disease might be the sustained
rence of mutations sometimes many decades before any diagnosis of normalization of blood cell counts without the need for phlebotomies,
disease109,110, highlight the likelihood that PV, or at least JAK2-mutated a normal or virtually normal bone marrow biopsy, and low-burden
CHIP, will be increasingly identified. This possibility raises the potential JAK2V617F variant allele frequency (perhaps even lower than 10%, rather
requirement for treatment to prevent the development of PV, or at least 1–5% or <1%). In addition, whether additional mutations in genes such
to minimize the associated enhanced risk of vascular events. These as TET2 and DNMT3A should also have a 50% reduction in variant allele
potential future interventions will require substantial efforts to identify frequency could be investigated. These aims all reflect important future
patients with high-risk CHIP and to develop a minimally toxic therapy. research priorities.
Historical data from a retrospective study in the 1970s, before As discussed above, agents targeting the hepcidin pathway could
current treatment algorithms evolved, showed that 50% of patients also prove useful adjuncts, as the burden associated with repeated
with PV died within 18 months, mainly due to thromboembolic events19. phlebotomy is high. Many agents are currently being evaluated, and
Over time, PV management has evolved to therapeutic phlebotomy, whether these will be used alone or as an adjunct to cytoreduction
aspirin and cytoreduction to reduce the risk of thromboembolic events, needs to be evaluated. Long-term risks and benefits of these treatments
with little evidence that this strategy will affect disease progression or will also need to be carefully scrutinized. We also need to investigate
overall survival, until recently (as discussed above). Targets for such what happens to patients when these agents are potentially discontin-
treatments were normalization of the surrogate of RBC mass, haema- ued and whether there will be a treatment escape mechanism. In the
tocrit and, controversially, the leukocyte and platelet counts, although past 5 years, interest has also focused on combination therapies; for
evidence for this approach has only just emerged. As discussed above, example, the combination of IFN and ruxolitinib. Also under investiga-
the MAJIC-PV study, published in 2023, suggested that therapy research tion is the combined use of cytoreductive agents for PV and a statin,
should now also focus on some form of molecular response (probably as statins have anti-inflammatory and positive vascular effects; such a
at least a 50% reduction in JAK2V617F variant allele frequency)116. Fur- combination has had an intriguing effect on the JAK2V617F variant allele
thermore, whilst in contemporary practice, cytoreductive treatments frequency177,178.
have only been utilized for high-risk PV, conventional lower-risk PV Patients with PV also face challenges in addition to vascular events,
is now increasingly recognized to have an adverse disease outcome. including uncertainty about future risks and the notable burden of

Table 4 | Investigational agents in clinical trials for the treatment of polycythaemia vera

Drug name Drug class Phase of ClinicalTrials.gov identifier Topline phase II results for agents in phase III trials
clinical
development

Rusfertide166 Hepcidin mimetic III NCT05210790 Response rate 60% vs 17% with placebo in double-blind, randomized
(peptide) withdrawal period (P = 0.002)
Sapablursen ASO against TMPRSS6 II NCT05143957 NA
Diversiran siRNA against TMPRSS6 I–II NCT05499013 NA
Bomedemstat Lysine demethylase-1 II NCT05558696 NA
inhibitor
Givinostat168–171 Histone deacetylase III NCT06093672 ORR >80% among 51 patients treated in a long-term rollover study or
inhibitor compassionate use; grade 3 AEs in 10% (median follow-up 2.8 years,
mean 4 years)
AE, adverse event; ASO, antisense oligonucleotide; NA, not applicable; ORR, overall response rate; siRNA, short interfering RNA.

Nature Reviews Disease Primers | (2025) 11:26 14

0123456789();:
Primer

Fig. 5 | Symptoms and quality of life in patients with

polycythaemia vera. Patients with polycythaemia
vera face a number of symptoms, quality of life and
• Headache
• Unexplained fatigue
disease-related effects. These include concerns
Visual disturbances
regarding complications of the disease such as
thrombosis, haemorrhage, symptoms and disease
transformation, but also wider effects, such as
concerns about disease inheritance. Patients might
also experience emotional and economical hardship as
QoL
• Emotional impact a consequence of their condition. QoL, quality of life.
• Economic impact
• Concerns about
familial predisposition

Aquagenic pruritus
Cardiac complications

Cancers Palpable
splenomegaly

Haemorrhage

Myelofibrosis
Erythromelalgia with fibrosis of
the bone marrow

Leukaemia
Thrombosis

symptoms; further progress in these areas is also required. In par- truly appreciate the risks of disease transformation, particularly to
ticular, support is required in identifying the causes of, and success- primary myelofibrosis and acute myeloid leukaemia, as well as under-
ful therapies for, fatigue and pruritus, which can be extraordinarily standing surrogate markers to indicate that treatments are reducing
limiting for patients. Fatigue, as discussed above, has major effects those risks. Furthermore, whilst a number of therapeutic options are
on many patients, and the aetiology of this particular highly preva- available for patients with PV, these are not curative and have consider-
lent symptom is unclear. Optimum strategies to improve this burden able adverse effects, especially the risk of non-melanoma skin cancers
need to be identified and made available. The second very troubling and, in the case of IFN therapy, mood and autoimmune disturbances.
but less prevalent symptom is pruritus; again, our understanding of Newer treatments, such as those targeting mutant JAK2 or pseudoki-
the pathophysiology of this symptom is not clear. Whilst our current nase domain-specific inhibitors should be evaluated and could offer
therapies are mostly successful in reducing symptoms, an improved the promise of less toxic and more targeted agents. For diseases such
understanding of the underlying cause(s) of these disease features as PV, efforts must also include the collection of detailed prospective
would help identify better therapy. biological and clinical data, which will facilitate better informed clinical
After more than 70 years from the first description of MPNs by pathways in the future.
William Dameshek179 and the clonal origin of the disease180,181, priorities
beyond the next 5–10 years include better stratification of patients to Published online: xx xx xxxx

Nature Reviews Disease Primers | (2025) 11:26 15

0123456789();:
Primer

References 34. Pedersen, K. M. et al. Smoking is associated with increased risk of myeloproliferative
1. Arber, D. A. et al. International Consensus Classification of Myeloid Neoplasms and neoplasms: a general population-based cohort study. Cancer Med. 7, 5796–5802 (2018).
Acute Leukemias: integrating morphologic, clinical, and genomic data. Blood 140, 35. Balandrán, J. C., Lasry, A. & Aifantis, I. The role of inflammation in the initiation and
1200–1228 (2022). progression of myeloid neoplasms. Blood Cancer Discov. 4, 254–266 (2023).
The 2022 ICC diagnostic criteria for myeloid neoplasms and acute leukaemias. Important review highlighting the role of inflammation both extrinsic and intrinsic to
2. James, C. et al. A unique clonal JAK2 mutation leading to constitutive signalling causes the mutant clone.
polycythaemia vera. Nature 434, 1144–1148 (2005). 36. Murphy, F. et al. Body size in relation to incidence of subtypes of haematological
Credited as the first description of the JAK2V617F mutation. malignancy in the prospective Million Women Study. Br. J. Cancer 108, 2390–2398 (2013).
3. Jaiswal, S. et al. Age-related clonal hematopoiesis associated with adverse outcomes. 37. Podoltsev, N. A. et al. Diet and risk of myeloproliferative neoplasms in older individuals
N. Engl. J. Med. 371, 2488–2498 (2014). from the NIH-AARP cohort. Cancer Epidemiol. Biomark. Prev. 29, 2343–2350 (2020).
4. Jaiswal, S. et al. Clonal hematopoiesis and risk of atherosclerotic cardiovascular disease. 38. Baxter, E. J. et al. Acquired mutation of the tyrosine kinase JAK2 in human
N. Engl. J. Med. 377, 111–121 (2017). myeloproliferative disorders. Lancet 365, 1054–1061 (2005).
5. Pearson, T. C. & Wetherley-Mein, G. Vascular occlusive episodes and venous haematocrit 39. Levine, R. L. et al. Activating mutation in the tyrosine kinase JAK2 in polycythemia vera,
in primary proliferative polycythaemia. Lancet 2, 1219–1222 (1978). essential thrombocythemia, and myeloid metaplasia with myelofibrosis. Cancer Cell 7,
6. Hultcrantz, M. et al. Risk for arterial and venous thrombosis in patients with 387–397 (2005).
myeloproliferative neoplasms: a population-based cohort study. Ann. Intern. Med. 168, 40. Kralovics, R. et al. A gain-of-function mutation of JAK2 in myeloproliferative disorders.
317–325 (2018). N. Engl. J. Med. 352, 1779–1790 (2005).
7. Ferrari, A. et al. Clinical outcomes under hydroxyurea treatment in polycythemia vera: 41. Scott, L. M. et al. JAK2 exon 12 mutations in polycythemia vera and idiopathic
a systematic review and meta-analysis. Haematologica 104, 2391–2399 (2019). erythrocytosis. N. Engl. J. Med. 356, 459–468 (2007).
8. Tefferi, A. & Barbui, T. Polycythemia vera: 2024 update on diagnosis, risk‐stratification, Credited as the first description of JAK2 exon 12 mutations.
and management. Am. J. Hematol. 98, 1465–1487 (2023). 42. Leroy, E. & Constantinescu, S. N. Rethinking JAK2 inhibition: towards novel strategies of
9. Barbui, T. et al. Philadelphia chromosome-negative classical myeloproliferative more specific and versatile janus kinase inhibition. Leukemia 31, 1023–1038 (2017).
neoplasms: revised management recommendations from European LeukemiaNet. 43. Constantinescu, S. N., Vainchenker, W., Levy, G. & Papadopoulos, N. Functional
Leukemia 32, 1057–1069 (2018). consequences of mutations in myeloproliferative neoplasms. HemaSphere 5, e578
10. Hultcrantz, M. et al. Patterns of survival among patients with myeloproliferative neoplasms (2021).
diagnosed in Sweden from 1973 to 2008: a population-based study. J. Clin. Oncol. 30, 44. Wingelhofer, B. et al. Implications of STAT3 and STAT5 signaling on gene regulation and
2995–3001 (2012). chromatin remodeling in hematopoietic cancer. Leukemia 32, 1713–1726 (2018).
11. Hultcrantz, M. et al. Risk and cause of death in patients diagnosed with 45. Liu, F. et al. JAK2V617F-mediated phosphorylation of PRMT5 downregulates its
myeloproliferative neoplasms in Sweden between 1973 and 2005: a population-based methyltransferase activity and promotes myeloproliferation. Cancer Cell 19, 283–294
study. J. Clin. Oncol. 33, 2288–2295 (2015). (2011).
12. Shallis, R. M. et al. Epidemiology of the classical myeloproliferative neoplasms: the four 46. Dawson, M. A. et al. JAK2 phosphorylates histone H3Y41 and excludes HP1α from
corners of an expansive and complex map. Blood Rev. 42, 100706 (2020). chromatin. Nature 461, 819–822 (2009).
13. Verstovsek, S. et al. Changes in the incidence and overall survival of patients 47. Bartalucci, N., Guglielmelli, P. & Vannucchi, A. M. Polycythemia vera: the current status
with myeloproliferative neoplasms between 2002 and 2016 in the United States. of preclinical models and therapeutic targets. Expert. Opin. Ther. Targets 24, 615–628
Leuk. Lymphoma 63, 694–702 (2022). (2020).
14. James, C., Ugo, V., Casadevall, N., Constantinescu, S. N. & Vainchenker, W. A JAK2 48. Lundberg, P. et al. Myeloproliferative neoplasms can be initiated from a single
mutation in myeloproliferative disorders: pathogenesis and therapeutic and scientific hematopoietic stem cell expressing JAK2-V617F. J. Exp. Med. 211, 2213–2230 (2014).
prospects. Trends Mol. Med. 11, 546–554 (2005). 49. Grisouard, J. et al. JAK2 exon 12 mutant mice display isolated erythrocytosis and changes
15. Titmarsh, G. J. et al. How common are myeloproliferative neoplasms? A systematic in iron metabolism favoring increased erythropoiesis. Blood 128, 839–851 (2016).
review and meta-analysis. Am. J. Hematol. 89, 581–587 (2014). 50. Passamonti, F. et al. Molecular and clinical features of the myeloproliferative neoplasm
16. Moulard, O. et al. Epidemiology of myelofibrosis, essential thrombocythemia, and associated with JAK2 exon 12 mutations. Blood 117, 2813–2816 (2011).
polycythemia vera in the European Union. Eur. J. Haematol. 92, 289–297 (2014). 51. Tefferi, A. et al. JAK2 exon 12 mutated polycythemia vera: Mayo-Careggi MPN Alliance
17. Hultcrantz, M. et al. Incidence of myeloproliferative neoplasms – trends by subgroup study of 33 consecutive cases and comparison with JAK2V617F mutated disease.
and age in a population-based study in Sweden. J. Intern. Med. 287, 448–454 (2020). Am. J. Hematol. 93, E93–E96 (2018).
18. Mehta, J., Wang, H., Iqbal, S. U. & Mesa, R. Epidemiology of myeloproliferative neoplasms 52. Spivak, J. L. Myeloproliferative neoplasms. N. Engl. J. Med. 376, 2168–2181 (2017).
in the United States. Leuk. Lymphoma 55, 595–600 (2014). 53. Mullally, A. et al. Distinct roles for long-term hematopoietic stem cells and erythroid
19. Chievitz, E. & Thiede, T. Complications and causes of death in polycythaemia vera. precursor cells in a murine model of Jak2V617F-mediated polycythemia vera. Blood 120,
Acta Med. Scand. 172, 513–523 (1962). 166–172 (2012).
20. Marchioli, R. et al. Vascular and neoplastic risk in a large cohort of patients with 54. Spivak, J. L. et al. Thrombopoietin is required for full phenotype expression in a
polycythemia vera. J. Clin. Oncol. 23, 2224–2232 (2005). JAK2V617F transgenic mouse model of polycythemia vera. PLoS ONE 15, e0232801
21. Barbui, T. et al. In contemporary patients with polycythemia vera, rates of thrombosis (2020).
and risk factors delineate a new clinical epidemiology. Blood 124, 3021–3023 (2014). 55. Moliterno, A. R., Hankins, W. D. & Spivak, J. L. Impaired expression of the thrombopoietin
22. Grinfeld, J. et al. Classification and personalized prognosis in myeloproliferative receptor by platelets from patients with polycythemia vera. N. Engl. J. Med. 338,
neoplasms. N. Engl. J. Med. 379, 1416–1430 (2018). 572–580 (1998).
23. Tefferi, A. et al. Mutation-enhanced international prognostic systems for essential 56. Tapper, W. et al. Genetic variation at MECOM, TERT, JAK2 and HBS1L-MYB predisposes to
thrombocythaemia and polycythaemia vera. Br. J. Haematol. 189, 291–302 (2020). myeloproliferative neoplasms. Nat. Commun. 6, 6691 (2015).
24. Luque Paz, D. et al. Leukemic evolution of polycythemia vera and essential 57. Lim, J., Ross, D. M., Brown, A. L., Scott, H. S. & Hahn, C. N. Germline genetic variants
thrombocythemia: genomic profiles predict time to transformation. Blood Adv. 4, that predispose to myeloproliferative neoplasms and hereditary myeloproliferative
4887–4897 (2020). phenotypes. Leuk. Res. 146, 107566 (2024).
25. Sobas, M. et al. Real-world study of children and young adults with myeloproliferative 58. Jones, A. V. et al. JAK2 haplotype is a major risk factor for the development of
neoplasms: identifying risks and unmet needs. Blood Adv. 6, 5171–5183 (2022). myeloproliferative neoplasms. Nat. Genet. 41, 446–449 (2009).
26. Goulart, H. et al. Myeloproliferative neoplasms in the adolescent and young adult 59. Olcaydu, D. et al. A common JAK2 haplotype confers susceptibility to myeloproliferative
population: a comprehensive review of the literature. Br. J. Haematol. 205, 48–60 (2024). neoplasms. Nat. Genet. 41, 450–454 (2009).
27. Heavner, K. et al. Working environment and myeloproliferative neoplasm: 60. Kilpivaara, O. et al. A germline JAK2 SNP is associated with predisposition to the
a population–based case-control study following a cluster investigation. Am. J. development of JAK2V617F-positive myeloproliferative neoplasms. Nat. Genet. 41, 455–459
Ind. Med. 58, 595–604 (2015). (2009).
28. Anderson, L. A. et al. Environmental, lifestyle, and familial/ethnic factors associated with 61. Li, S. L., Zhang, P. J., Sun, G. X. & Lu, Z. J. The JAK2 46/1 haplotype (GGCC) in
myeloproliferative neoplasms. Am. J. Hematol. 87, 175–182 (2012). myeloproliferative neoplasms and splanchnic vein thrombosis: a pooled analysis of
A useful summary of the epidemiology of MPNs. 26 observational studies. Ann. Hematol. 93, 1845–1852 (2014).
29. Merk, K. et al. The incidence of cancer among blood donors. Int. J. Epidemiol. 19, 62. Tefferi, A. et al. Targeted deep sequencing in polycythemia vera and essential
505–509 (1990). thrombocythemia. Blood Adv. 1, 21–30 (2016).
30. Najean, Y., Rain, J. D. & Billotey, C. Epidemiological data in polycythaemia vera: a study of 63. Ortmann, C. A. et al. Effect of mutation order on myeloproliferative neoplasms.
842 cases. Hematol. Cell Ther. 40, 159–165 (1998). N. Engl. J. Med. 372, 601–612 (2015).
31. Edgren, G. et al. Blood donation and risk of polycythemia vera. Transfusion 56, 1622–1627 64. Chen, E. et al. Distinct effects of concomitant Jak2V617F expression and Tet2 loss in mice
(2016). combine to promote disease progression in myeloproliferative neoplasms. Blood 125,
32. Duncombe, A. S. et al. Modifiable lifestyle and medical risk factors associated with 327–335 (2015).
myeloproliferative neoplasms. Hemasphere 4, e327 (2020). 65. Knudsen, T. A. et al. Genomic profiling of a randomized trial of interferon-α vs
33. McMullin, M. F. & Anderson, L. A. Aetiology of myeloproliferative neoplasms. Cancers 12, hydroxyurea in MPN reveals mutation-specific responses. Blood Adv. 6, 2107–2119
1810 (2020). (2022).

Nature Reviews Disease Primers | (2025) 11:26 16

0123456789();:
Primer

66. Abu-Zeinah, G., Erdos, K., Lee, N. Jr., Silver, R. T. & Scandura, J. M. DNMT3A mutations do 98. Galtier, J. et al. Role of red cell mass evaluation in myeloproliferative neoplasms with
not affect treatment response or outcomes for patients with polycythemia vera treated splanchnic vein thrombosis and normal hemoglobin value: a study of the France
with interferon alpha [abstract]. Blood 144, 3190 (2024). Intergroupe des Syndromes myeloproliferatifs. Haematologica 109, 1989–1993 (2024).
67. Shimizu, T. et al. Loss of Ezh2 synergizes with JAK2-V617F in initiating myeloproliferative 99. Barbui, T. et al. Masked polycythemia Vera (mPV): results of an international study.
neoplasms and promoting myelofibrosis. J. Exp. Med. 213, 1479–1496 (2016). Am. J. Hematol. 89, 52–54 (2014).
68. Usart, M. et al. Loss of Dnmt3a increases self-renewal and resistance to pegIFN-α in 100. Barbui, T., Thiele, J., Vannucchi, A. M. & Tefferi, A. Rethinking the diagnostic criteria of
JAK2-V617F-positive myeloproliferative neoplasms. Blood 143, 2490–2503 (2024). polycythemia vera. Leukemia 28, 1191–1195 (2014).
69. Theocharides, A. et al. Leukemic blasts in transformed JAK2-V617F-positive 101. Barbui, T. et al. Initial bone marrow reticulin fibrosis in polycythemia vera exerts an
myeloproliferative disorders are frequently negative for the JAK2-V617F mutation. Blood impact on clinical outcome. Blood 119, 2239–2241 (2012).
110, 375–379 (2007). 102. Wilkins, B. S. et al. Bone marrow pathology in essential thrombocythemia: interobserver
70. Calabresi, L. et al. Clonal dynamics and copy number variants by single‐cell analysis reliability and utility for identifying disease subtypes. Blood 111, 60–70 (2008).
in leukemic evolution of myeloproliferative neoplasms. Am. J. Hematol. 98, 1520–1532 103. Madelung, A. B. et al. WHO-defined classification of myeloproliferative neoplasms:
(2023). morphological reproducibility and clinical correlations – the Danish experience.
71. Tefferi, A. et al. Survival and prognosis among 1545 patients with contemporary Am. J. Hematol. 88, 1012–1016 (2013).
polycythemia vera: an international study. Leukemia 27, 1874–1881 (2013). 104. Ryou, H. et al. Continuous Indexing of Fibrosis (CIF): improving the assessment and
Contemporary study of mortality of a cohort of patients with PV. classification of MPN patients. Leukemia 37, 348–358 (2023).
72. Tang, G. et al. Characteristics and clinical significance of cytogenetic abnormalities in 105. Scott, L. M. et al. The V617F JAK2 mutation is uncommon in cancers and in myeloid
polycythemia vera. Haematologica 102, 1511–1518 (2017). malignancies other than the classic myeloproliferative disorders. Blood 106, 2920–2921
73. Gangat, N. et al. Cytogenetic studies at diagnosis in polycythemia vera: clinical and (2005).
JAK2V617F allele burden correlates. Eur. J. Haematol. 80, 197–200 (2008). 106. Ma, W. et al. Mutation profile of JAK2 transcripts in patients with chronic
74. Fisher, D. A. C., Fowles, J. S., Zhou, A. & Oh, S. T. Inflammatory pathophysiology as a myeloproliferative neoplasias. J. Mol. Diagn. 11, 49–53 (2009).
contributor to myeloproliferative neoplasms. Front. Immunol. 12, 683401 (2021). 107. McMullin, M. F. & Cario, H. LNK mutations and myeloproliferative disorders. Am. J. Hematol.
75. Barbui, T. et al. Inflammation and thrombosis in essential thrombocythemia and 91, 248–251 (2016).
polycythemia vera: different role of C-reactive protein and Pentraxin 3. Haematologica 108. Milosevic Feenstra, J. D. et al. Whole-exome sequencing identifies novel MPL and JAK2
96, 315–318 (2011). mutations in triple-negative myeloproliferative neoplasms. Blood 127, 325–332 (2016).
76. Vaidya, R. et al. Plasma cytokines in polycythemia vera: phenotypic correlates, prognostic 109. Williams, N. et al. Life histories of myeloproliferative neoplasms inferred from
relevance, and comparison with myelofibrosis. Am. J. Hematol. 87, 1003–1005 (2012). phylogenies. Nature 602, 162–168 (2022).
77. Tefferi, A. et al. Circulating interleukin (IL)-8, IL-2R, IL-12, and IL-15 levels are Use of mathematical modelling to illustrate that a driver JAK2 mutation could arise
independently prognostic in primary myelofibrosis: a comprehensive cytokine profiling several decades before the onset of disease.
study. J. Clin. Oncol. 29, 1356–1363 (2011). 110. How, J., Garcia, J. S. & Mullally, A. Biology and therapeutic targeting of molecular
78. Laranjeira, A. B. A. et al. In vivo ablation of NFκB cascade effectors alleviates disease mechanisms in MPNs. Blood 141, 1922–1933 (2023).
burden in myeloproliferative neoplasms. Blood J. 143, 2414–2424 (2024). 111. Vannucchi, A. M. et al. Clinical profile of homozygous JAK2 617V>F mutation in patients
79. Rai, S. et al. IL-1β promotes MPN disease initiation by favoring early clonal expansion of with polycythemia vera or essential thrombocythemia. Blood 110, 840–846 (2007).
JAK2-mutant hematopoietic stem cells. Blood Adv. 8, 1234–1249 (2024). 112. Passamonti, F. et al. A prospective study of 338 patients with polycythemia vera: the
80. Rai, S. et al. Inhibition of interleukin-1β reduces myelofibrosis and osteosclerosis in impact of JAK2 (V617F) allele burden and leukocytosis on fibrotic or leukemic disease
mice with JAK2-V617F driven myeloproliferative neoplasm. Nat. Commun. 13, 5346 transformation and vascular complications. Leukemia 24, 1574–1579 (2010).
(2022). 113. Gisslinger, H. et al. Ropeginterferon alfa-2b versus standard therapy for polycythaemia
81. Verstovsek, S. et al. Safety and efficacy of INCB018424, a JAK1 and JAK2 inhibitor, in vera (PROUD-PV and CONTINUATION-PV): a randomised, non-inferiority, phase 3 trial and
myelofibrosis. N. Engl. J. Med. 363, 1117–1127 (2010). its extension study. Lancet Haematol. 7, e196–e208 (2020).
First description of the use of ruxolitinib in myelofibrosis. An important study (PROUD-PV) comparing ropeginterferon alfa-2b with
82. Verstovsek, S. et al. A phase 2 study of ruxolitinib, an oral JAK1 and JAK2 inhibitor, hydroxycarbamide.
in patients with advanced polycythemia vera who are refractory or intolerant to 114. Kiladjian, J. J. et al. Pegylated interferon-alfa-2a induces complete hematological and
hydroxyurea. Cancer 120, 513–520 (2014). molecular responses with low toxicity in polycythemia vera. Blood 112, 3065–3072
First study of ruxolitinib in PV. (2008).
83. Ramanathan, G. et al. Cigarette smoke stimulates clonal expansion of Jak2V617F and Tet2−/− 115. Kiladjian, J. J. et al. Long-term efficacy and safety of ruxolitinib versus best available
cells. Front. Oncol. 13, 1210528 (2023). therapy in polycythaemia vera (RESPONSE): 5-year follow up of a phase 3 study.
84. Pratt, J. J. & Khan, K. S. Non-anaemic iron deficiency – a disease looking for recognition of Lancet Haematol. 7, e226–e237 (2020).
diagnosis: a systematic review. Eur. J. Haematol. 96, 618–628 (2016). Long-term follow-up of the RESPONSE study in PV.
85. Verstovsek, S. et al. Markers of iron deficiency in patients with polycythemia vera 116. Harrison, C. N. et al. Ruxolitinib versus best available therapy for polycythemia vera
receiving ruxolitinib or best available therapy. Leuk. Res. 56, 52–59 (2017). intolerant or resistant to hydroxycarbamide in a randomized trial. J. Clin. Oncol. 41,
86. Song, J., Kim, S. J., Min, G., Lim, Y. & Prchal, J. T. Iron deficiency correction in 3534–3544 (2023).
myeloproliferative neoplasms reduces thrombosis risk via decreased P-selectin. Blood MAJIC-PV demonstrating the long-term benefit for ruxolitinib compared with best
144, 1758–1758 (2024). available therapy with improved event-free survival in patients treated with ruxolitinib,
87. Ganz, T. Anemia of inflammation. N. Engl. J. Med. 381, 1148–1157 (2019). control of blood count and 50% reduction in JAK2V617F variant allele frequency.
88. Kautz, L. et al. Identification of erythroferrone as an erythroid regulator of iron 117. Guglielmelli, P. et al. Clinical impact of mutated JAK2 allele burden reduction in
metabolism. Nat. Genet. 46, 678–684 (2014). polycythemia vera and essential thrombocythemia. Am. J. Hematol. 99, 1550–1559 (2024).
89. Handa, S., Ginzburg, Y., Hoffman, R. & Kremyanskaya, M. Hepcidin mimetics in polycythemia Study demonstrating that reduction of JAK2V617F variant allele frequency correlates with
vera: resolving the irony of iron deficiency and erythrocytosis. Curr. Opin. Hematol. 30, lower risk of post-PV myelofibrosis.
45–52 (2023). 118. Delhommeau, F. et al. Oncogenic mechanisms in myeloproliferative disorders.
90. Casu, C. et al. Minihepcidin peptides as disease modifiers in mice affected by Cell Mol. Life Sci. 63, 2939–2953 (2006).
β-thalassemia and polycythemia vera. Blood 128, 265–276 (2016). 119. Pasquer, H. et al. Distinct clinico-molecular arterial and venous thrombosis scores for
91. Andrikovics, H. et al. HFE C282Y mutation as a genetic modifier influencing disease myeloproliferative neoplasms risk stratification. Leukemia 38, 326–339 (2024).
susceptibility for chronic myeloproliferative disease. Cancer Epidemiol. Biomark. Prev. 120. Santos, F. P. S. et al. Prognostic impact of RAS-pathway mutations in patients with
18, 929–934 (2009). myelofibrosis. Leukemia 34, 799–810 (2020).
92. Stetka, J. et al. Iron is a modifier of the phenotypes of JAK2-mutant myeloproliferative 121. O’Sullivan, J. M. et al. RAS-pathway mutations are common in patients with ruxolitinib
neoplasms. Blood 141, 2127–2140 (2023). refractory/intolerant myelofibrosis: molecular analysis of the PAC203 cohort. Leukemia
93. van Genderen, P. J. & Michiels, J. J. Erythromelalgia: a pathognomonic microvascular 37, 2497–2501 (2023).
thrombotic complication in essential thrombocythemia and polycythemia vera. 122. Quintás-Cardama, A. et al. Molecular analysis of patients with polycythemia vera or
Semin. Thromb. Hemost. 23, 357–363 (1997). essential thrombocythemia receiving pegylated interferon alpha-2a. Blood 122, 893–901
94. Kiladjian, J. J. & Cassinat, B. Myeloproliferative neoplasms and splanchnic vein thrombosis: (2013).
contemporary diagnostic and therapeutic strategies. Am. J. Hematol. 98, 794–800 (2023). 123. Bewersdorf, J. P. et al. Moving toward disease modification in polycythemia vera. Blood
95. Khoury, J. D. et al. The 5th edition of the World Health Organization classification of 142, 1859–1870 (2023).
haematolymphoid tumours: myeloid and histiocytic/dendritic neoplasms. Leukemia 36, 124. McMullin, M. F. & Harrison, C. N. How I treat low-risk polycythemia vera patients who
1703–1719 (2022). require cytoreduction. Blood https://doi.org/10.1182/blood.2023022418 (2024).
The 2022 WHO criteria for diagnosis of haematolymphoid disorders. 125. Gerds, A. T. et al. Myeloproliferative neoplasms, version 3.2022, NCCN clinical practice
96. Thiele, J. et al. The International Consensus Classification of Myeloid Neoplasms and guidelines in oncology. J. Natl Compr. Cancer Netw. 20, 1033–1062 (2022).
Acute Leukemias: myeloproliferative neoplasms. Am. J. Hematol. 98, 166–179 (2023). 126. Marchetti, M. et al. Appropriate management of polycythaemia vera with cytoreductive
97. Maslah, N. et al. Masked polycythemia vera: analysis of a single center cohort of 2480 drug therapy: European LeukemiaNet 2021 recommendations. Lancet Haematol. 9,
red cell masses. Haematologica 105, e95–e97 (2020). e301–e311 (2022).

Nature Reviews Disease Primers | (2025) 11:26 17

0123456789();:
Primer

127. Landolfi, R. et al. Leukocytosis as a major thrombotic risk factor in patients with 159. Passamonti, F. et al. Ruxolitinib for the treatment of inadequately controlled
polycythemia vera. Blood 109, 2446–2452 (2007). polycythaemia vera without splenomegaly (RESPONSE-2): a randomised, open-label,
128. Barbui, T. et al. White blood cell counts and thrombosis in polycythemia vera: a subanalysis phase 3b study. Lancet Oncol. 18, 88–99 (2017).
of the CYTO-PV study. Blood 126, 560–561 (2015). Second registration trial of ruxolitinib in the treatment of PV.
129. Ronner, L. et al. Persistent leukocytosis in polycythemia vera is associated with disease 160. Masciulli, A., Ferrari, A., Carobbio, A., Ghirardi, A. & Barbui, T. Ruxolitinib for the
evolution but not thrombosis. Blood 135, 1696–1703 (2020). prevention of thrombosis in polycythemia vera: a systematic review and meta-analysis.
130. Gerds, A. T. et al. Association between elevated white blood cell counts and thrombotic Blood Adv. 4, 380–386 (2020).
events in polycythemia vera: analysis from REVEAL. Blood 143, 1646–1655 (2024). 161. Vannucchi, A. M. et al. Ruxolitinib reduces JAK2 p.V617F allele burden in patients
131. Carobbio, A. et al. Neutrophil-to-lymphocyte ratio is a novel predictor of venous with polycythemia vera enrolled in the RESPONSE study. Ann. Hematol. 96, 1113–1120
thrombosis in polycythemia vera. Blood Cancer J. 12, 28 (2022). (2017).
132. Larsen, M. K. et al. Neutrophil-to-lymphocyte ratio and all-cause mortality with and without 162. Koschmieder, S. et al. Firstline treatment with ruxolitinib versus best available therapy in
myeloproliferative neoplasms – a Danish longitudinal study. Blood Cancer J. 14, 28 (2024). patients with polycythemia vera: pre-specified interim analysis of the randomized phase
133. Moliterno, A. R., Kaizer, H. & Reeves, B. N. JAK2V617F allele burden in polycythemia vera: 2b Ruxobeat clinical trial of the German Study Group for Myeloproliferative Neoplasms
burden of proof. Blood 141, 1934–1942 (2023). (GSG-MPN) [abstract]. Blood 142, 619 (2023).
134. Vannucchi, A. M. et al. Prospective identification of high-risk polycythemia vera patients 163. Passamonti, F. et al. Ruxolitinib versus best available therapy in inadequately controlled
based on JAK2V617F allele burden. Leukemia 21, 1952–1959 (2007). polycythaemia vera without splenomegaly (RESPONSE-2): 5-year follow up of a
135. Guglielmelli, P. et al. JAK2V617F variant allele frequency >50% identifies patients with randomised, phase 3b study. Lancet Haematol. 9, e480–e492 (2022).
polycythemia vera at high risk for venous thrombosis. Blood Cancer J. 11, 199 (2021). 164. Ginzburg, Y. Z. et al. Dysregulated iron metabolism in polycythemia vera: etiology and
136. Marchioli, R. et al. Cardiovascular events and intensity of treatment in polycythemia vera. consequences. Leukemia 32, 2105–2116 (2018).
N. Engl. J. Med. 368, 22–33 (2013). 165. Casu, C., Nemeth, E. & Rivella, S. Hepcidin agonists as therapeutic tools. Blood 131,
137. Landolfi, R. et al. Efficacy and safety of low-dose aspirin in polycythemia vera. N. Engl. J. Med. 1790–1794 (2018).
350, 114–124 (2004). 166. Kremyanskaya, M. et al. Rusfertide, a hepcidin mimetic, for control of erythrocytosis in
138. Barbui, T. et al. A reappraisal of the benefit-risk profile of Hydroxyurea in polycythemia polycythemia vera. N. Engl. J. Med. 390, 723–735 (2024).
vera: a propensity-matched study. Am. J. Hematol. 92, 1131–1136 (2017). First description of the use of rusfertide, a hepcidin mimetic, in the management of PV.
139. Podoltsev, N. A. et al. The impact of phlebotomy and hydroxyurea on survival and risk of 167. Guerini, V. et al. The histone deacetylase inhibitor ITF2357 selectively targets cells
thrombosis among older patients with polycythemia vera. Blood Adv. 2, 2681–2690 (2018). bearing mutated JAK2V617F. Leukemia 22, 740–747 (2008).
140. Barbui, T. et al. No correlation of intensity of phlebotomy regimen with risk of thrombosis 168. Rambaldi, A. et al. A pilot study of the histone-deacetylase inhibitor givinostat in patients
in polycythemia vera: evidence from ECLAP and CYTO-PV clinical trials. Haematologica with JAK2V617F positive chronic myeloproliferative neoplasms. Br. J. Haematol. 150,
102, e219–e221 (2017). 446–455 (2010).
141. Alvarez-Larran, A. et al. Risk of thrombosis according to need of phlebotomies in patients 169. Rambaldi, A. et al. Safety and efficacy of the maximum tolerated dose of givinostat in
with polycythemia vera treated with hydroxyurea. Haematologica 102, 103–109 (2017). polycythemia vera: a two-part phase Ib/II study. Leukemia 34, 2234–2237 (2020).
142. Finazzi, G. et al. Acute leukemia in polycythemia vera: an analysis of 1638 patients 170. Finazzi, G. et al. A phase II study of givinostat in combination with hydroxycarbamide
enrolled in a prospective observational study. Blood 105, 2664–2670 (2005). in patients with polycythaemia vera unresponsive to hydroxycarbamide monotherapy.
143. Kiladjian, J.-J., Chevret, S., Dosquet, C., Chomienne, C. & Rain, J.-D. Treatment of Br. J. Haematol. 161, 688–694 (2013).
polycythemia vera with hydroxyurea and pipobroman: final results of a randomized trial 171. Rambaldi, A. et al. Long-term safety and efficacy of givinostat in polycythemia vera:
initiated in 1980. J. Clin. Oncol. 29, 3907–3913 (2011). 4-year mean follow up of three phase 1/2 studies and a compassionate use program.
144. Landtblom, A. R. et al. Second malignancies in patients with myeloproliferative neoplasms: Blood Cancer J. 11, 53 (2021).
a population-based cohort study of 9379 patients. Leukemia 32, 2203–2210 (2018). 172. Mesa, R. A. et al. The burden of fatigue and quality of life in myeloproliferative disorders
145. Barbui, T. et al. Second cancer in Philadelphia negative myeloproliferative neoplasms (MPDs): an international Internet-based survey of 1179 MPD patients. Cancer 109, 68–76
(MPN-K). A nested case-control study. Leukemia 33, 1996–2005 (2019). (2007).
146. Vachhani, P. et al. Interferons in the treatment of myeloproliferative neoplasms. 173. Mesa, R. A. et al. The Myelofibrosis Symptom Assessment Form (MFSAF): an
Ther. Adv. Hematol. 15, 20406207241229588 (2024). evidence-based brief inventory to measure quality of life and symptomatic response to
147. Gisslinger, H. et al. Event-free survival in patients with polycythemia vera treated with treatment in myelofibrosis. Leuk. Res. 33, 1199–1203 (2009).
ropeginterferon alfa-2b versus best available treatment. Leukemia 37, 2129–2132 (2023). 174. Scherber, R. et al. The myeloproliferative neoplasm symptom assessment form
Demonstration that interferons lead to superior event-free survival in PV. (MPN-SAF): international prospective validation and reliability trial in 402 patients. Blood
148. Barbui, T. et al. Ropeginterferon alfa-2b versus phlebotomy in low-risk patients with 118, 401–408 (2011).
polycythaemia vera (Low-PV study): a multicentre, randomised phase 2 trial. Lancet The development and validation of an international tool to measure patient symptoms.
Haematol. 8, e175–e184 (2021). 175. Mesa, R. A. et al. NCCN guidelines insights: myeloproliferative neoplasms, version
The first study specifically evaluating interferon in low-risk patients with PV. 2.2018. J. Natl Compr. Cancer Netw. 15, 1193–1207 (2017).
149. Barbui, T. et al. Ropeginterferon phase 2 randomized study in low-risk polycythemia 176. Mazza, G. L. et al. Symptom burden and quality of life in patients with high-risk essential
vera: 5-year drug survival and efficacy outcomes. Ann. Hematol. 103, 437–442 (2024). thrombocythaemia and polycythaemia vera receiving hydroxyurea or pegylated
150. Mascarenhas, J. et al. A randomized phase 3 trial of interferon-α vs hydroxyurea in interferon alfa-2a: a post-hoc analysis of the MPN-RC 111 and 112 trials. Lancet Haematol.
polycythemia vera and essential thrombocythemia. Blood 139, 2931–2941 (2022). 9, e38–e48 (2022).
151. Abu-Zeinah, G. et al. Interferon-alpha for treating polycythemia vera yields improved 177. Sørensen, A. L. et al. Combination therapy with ruxolitinib and pegylated interferon
myelofibrosis-free and overall survival. Leukemia 35, 2592–2601 (2021). alfa-2a in newly diagnosed patients with polycythemia vera. Blood Adv. 8, 5416–5425
Long-term study demonstrating benefits of interferon on myelofibrosis-free survival. (2024).
152. Kiladjian, J. J. et al. Clonal analysis of erythroid progenitors suggests that pegylated 178. Hansen, I. D. et al. Statins enhance the efficacy of pegylated interferon-alpha2 in
interferon α-2a treatment targets JAK2V617F clones without affecting TET2 mutant cells. patients with pH-negative chronic myeloproliferative neoplasms. results from a Danish
Leukemia 24, 1519–1523 (2010). single-institution cohort study [abstract]. Blood 144, 3184 (2024).
153. Daltro De Oliveira, R. et al. Interferon-alpha (IFN) therapy discontinuation is feasible in 179. Dameshek, W. Some speculations on the myeloproliferative syndromes. Blood 6,
myeloproliferative neoplasm (MPN) patients with complete hematological remission 372–375 (1951).
[abstract]. Blood 136, 35–36 (2020). Seminal description by Dameshek of the family of disorders where he first used the
Important demonstration that interferon can be discontinued in a proportion of term myeloproliferative.
patients with PV. 180. Prchal, J. F. & Axelrad, A. A. Letter: bone-marrow responses in polycythemia vera.
154. Masarova, L. et al. Pegylated interferon alfa-2a in patients with essential N. Engl. J. Med. 290, 1382 (1974).
thrombocythaemia or polycythaemia vera: a post-hoc, median 83 month follow-up of an 181. Adamson, J. W., Fialkow, P. J., Murphy, S., Prchal, J. F. & Steinmann, L. Polycythemia vera:
open-label, phase 2 trial. Lancet Haematol. 4, e165–e175 (2017). stem-cell and probable clonal origin of the disease. N. Engl. J. Med. 295, 913–916 (1976).
155. Kiladjian, J. J. et al. Long-term outcomes of polycythemia vera patients treated with Demonstration of clonal origin of PV using X-chromosome inactivation patterns.
ropeginterferon alfa-2b. Leukemia 36, 1408–1411 (2022). 182. Barbui, T. et al. Philadelphia-negative classical myeloproliferative neoplasms: critical
156. Barosi, G. et al. A unified definition of clinical resistance and intolerance to concepts and management recommendations from European LeukemiaNet. J. Clin. Oncol.
hydroxycarbamide in polycythaemia vera and primary myelofibrosis: results of a 29, 761–770 (2011).
European LeukemiaNet (ELN) consensus process. Br. J. Haematol. 148, 961–963 (2010). 183. Gu, W. et al. Prediction of thrombosis in polycythemia vera: development and validation
157. Alvarez-Larran, A. et al. Assessment and prognostic value of the European LeukemiaNet of a multiple factor-based prognostic score system. Res. Pract. Thromb. Haemost. 7,
criteria for clinicohematologic response, resistance, and intolerance to hydroxyurea in 100132 (2023).
polycythemia vera. Blood 119, 1363–1369 (2012). 184. Yacoub, A. et al. Pegylated interferon alfa-2a for polycythemia vera or essential
158. Vannucchi, A. M. et al. Ruxolitinib versus standard therapy for the treatment of thrombocythemia resistant or intolerant to hydroxyurea. Blood 134, 1498–1509 (2019).
polycythemia vera. N. Engl. J. Med. 372, 426–435 (2015). 185. Samuelsson, J. et al. A phase II trial of pegylated interferon alpha-2b therapy for
Seminal study leading to the approval of ruxolitinib for the treatment of PV polycythemia vera and essential thrombocythemia: feasibility, clinical and biologic
(the RESPONSE trial). effects, and impact on quality of life. Cancer 106, 2397–2405 (2006).

Nature Reviews Disease Primers | (2025) 11:26 18

0123456789();:
Primer

186. Knudsen, T. A. Three-year analysis of the DALIAH trial – a randomized controlled phase III Keros, Pharma Essentia, Jubilant, Morphic, Novartis, Blueprint, Ono, Raythera and Cogent.
clinical trial comparing recombinant interferon alpha-2 vs. hydroxyurea in patients with J.-J.K. discloses honoraria/consulting fees from Novartis, GSK, Abbvie, BMS, Incyte, AOP Health
myeloproliferative neoplasms [abstract S1609]. Hemasphere 3, 741–742 (2019). and PharmaEssentia. J.M. discloses research funding from Incyte, Novartis, BMS, CTI/SOBI,
187. Gisslinger, H. et al. Ropeginterferon alfa-2b, a novel IFNα-2b, induces high response rates Abbvie, Geron, Kartos, Karyopharm AJAX, Italfarmaco Spa, Disc and PharmaEssentia; and
with low toxicity in patients with polycythemia vera. Blood 126, 1762–1769 (2015). consulting fees from Incyte, Novartis, BMS, Geron, Karyopharm, Kartos, GSK, PharmaEssentia,
188. Barbui, T. et al. Ropeginterferon versus standard therapy for low-risk patients with Italfarmaco Spa, Abbvie, Roche, Merck, Pfizer, Galecto, MorphoSys, Disc, Keros and
polycythemia vera. NEJM Evid. 2, EVIDoa2200335 (2023). Sumitomo. M.F.M. discloses research support from BMS and AOP; and honorarium/consulting
189. Robinson, S., Ragheb, M. & Harrison, C. How I treat myeloproliferative neoplasms in fees from Novartis, GSK, Incyte, BMS and AOP. R.M. discloses research support from Abbvie,
pregnancy. Blood 143, 777–785 (2024). Blueprint, BMS, CTI, Genentech, Incyte, Morphosys and Sierra; and honoraria/consulting fees
from Abbvie, Blueprint, BMS, CTI, Genentech, Geron, GSK, Incyte, Novartis, Sierra, Sierra
Author contributions Oncology and Telios. A.M.V. discloses honoraria/consulting fees from Incyte, Novartis, AOP,
Introduction (T.B.); Epidemiology (M.F.M.); Mechanisms/pathophysiology (A.M.V.); Diagnosis, Italfarmaco, BMS, GSK, Abbvie, Blueprint and Ionis Disc.
screening and prevention (J.-J.K.); Management (P.B. and J.M.); Quality of life (R.M.); Outlook
(C.N.H.); overview of the Primer (C.N.H. and A.M.V.). Additional information
Peer review information Nature Reviews Disease Primers thanks H. Hasselbalch; N. Komatsu,
Competing interests who co-reviewed with Y. Edahiro; H. L. Pahl; and J. Prchal, who co-reviewed with J. Song, for
C.N.H. discloses research support from Celgene (BMS), Constellation, GSK and Novartis; their contribution to the peer review of this work.
and honoraria/consulting fees from Abbvie, AOP, BMS, CTI, IMAGO, Incyte, Novartis, Galacteo,
Geron, GSK, Janssen, Keros, MSD, SOBI and Morphosys. T.B. discloses research support ftom Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in
GSK and AOP; and honoraria/consulting fees from AOP, Italfarmaco, Ionis and Novartis. P.B. published maps and institutional affiliations.
discloses research support from Incyte, BMS, CTI, Morphosys, Sumitomo, Karyopharm, Kartos,
Telios, Ionis, Disc, Ajax, Geron, Janssen, Blueprint and Cogent; and honoraria/consulting fees © Crown 2025
from Incyte, BMS, CTI, GSK, Abbvie, Morphosys, Sumitomo, Karyopharm, Ionis, Disc, Geron,

Nature Reviews Disease Primers | (2025) 11:26 19

0123456789();: