Impaired iron balance and erythrocytosis: a complex relationship

Abstract

Erythropoiesis and iron metabolism are closely interconnected, under both physiological and pathological conditions. Polycythemia vera (PV) is a myeloproliferative neoplasm (MPN) caused by a Janus kinase-2 (JAK2) mutation, resulting in uncontrolled red blood cell production and elevated hemoglobin and hematocrit levels. In PV, iron deficiency is common and may be due to several factors, including chronic gastrointestinal bleeding, chronic inflammation, dysregulated hepcidin signaling, and therapeutic phlebotomy. Many patients exhibit low serum ferritin and low to normal hepcidin levels despite erythroid proliferation, suggesting a maladaptive iron-restricted state. This functional iron deficiency may limit erythropoiesis to some extent, but also contributes to burdensome symptoms such as fatigue, cognitive impairment, and restless leg syndrome. Novel therapeutic approaches, including hepcidin mimetics or ferroportin inhibitors, aim to restore iron homeostasis, improving quality of life and potentially reducing the need for cytoreductive therapy (drugs used to treat MPNs by reducing blood cell production) in low-risk PV patients. In secondary forms of erythrocytosis (both congenital and acquired), iron homeostasis has been less often investigated. However, exploring it may be of great interest since in these affections, iron overload could act as a hidden driver of erythrocytosis by stimulating erythroblast proliferation. Hereditary hemochromatosis (HH) is a well-known cause of iron overload. While its association with erythrocytosis has been a subject of interest, it remains incompletely understood. In this review, we will explore the complex relationship between iron (deficiency or overload, including HH) and erythrocytosis (including PV), discussing underlying mechanisms and potential therapeutic applications.

Introduction

At steady state, the human body contains approximately 250 billion erythroblasts in the bone marrow and 25 trillion red blood cells (RBC) in circulation. Erythropoiesis produces around two million RBC per second in humans, making it one of the most active proliferative systems in the body. It consumes nearly 80% of circulating iron. Both erythropoiesis and iron homeostasis require finely tuned regulation to ensure that RBC production and iron availability are continuously adapted to physiological needs.

Several factors are essential for effective erythropoiesis, including erythropoietin (EPO) and stem cell factor (SCF). Erythroblasts also express high levels of transferrin receptor type 1 (TfR1 or CD71), to uptake the iron necessary for hemoglobin synthesis. Dysregulation of erythropoiesis can result in either decreased RBC production (anemia) or excessive production (erythrocytosis). Polycythemia Vera (PV) is the leading cause of primary erythrocytosis.

Systemic iron distribution is finely regulated by the liver hormone hepcidin, which inhibits ferroportin, the only known iron export protein. Disruption of iron homeostasis can lead to iron deficiency or iron overload. Hereditary hemochromatosis (HH) is a genetic disorder characterized by iron overload due to low hepcidin levels.

Although the link between iron and erythropoiesis is well established, the mechanisms connecting iron regulation and erythrocytosis remain incompletely understood. In this review, we will explore the current comprehension of these interactions, focusing on iron (deficiency or overload, including HH) and erythrocytosis (including PV), and discuss possible therapeutic implications.

Polycythemia Vera (PV)

PV is one of the main Philadelphia chromosome-negative myeloproliferative neoplasms (MPNs), a heterogeneous group of clonal hematopoietic stem cell disorders.

PV is characterized by erythrocytosis (increased RBC count, hemoglobin and hematocrit) and by trilineage hyperplasia, particularly erythroid, in the bone marrow. Common symptoms include aquagenic pruritus, microvascular symptoms, fatigue, and splenomegaly. PV is associated with a significantly increased risk of arterial and venous thrombosis and can progress to myelofibrosis or acute myeloid leukemia (both associated with poor prognosis) [1].

Virtually all PV cases are driven by mutations in JAK2 (Janus Kinase 2), a non-receptor tyrosine kinase involved in erythropoietin receptor (EPO-R) signaling. Around 98% of PV patients harbor the JAK2 V617F mutation [2], while ~2% carry mutations in exon 12 of JAK2 [3, 4]. Other rare JAK2 mutations have also been described [5]. The remaining cases may involve other mutations affecting JAK/STAT (signal transducer and activator of transcription) signaling, either known, such as LNK mutations (lymphocyte adapter protein or SH2B3, a negative regulator of JAK–STAT signaling) [6], or yet to be identified. These mutations result in constitutive EPO-independent activation of the JAK/STAT pathway, promoting erythroid precursor survival, proliferation, and differentiation.

Some differences can be observed between JAK2^V617F and JAK2^exon12 mutations:

• The JAK2^V617F mutation can be found in PV but also in other MPNs (essential thrombocythemia (ET) and primary myelofibrosis (MF)) and may be associated with thrombocytosis and leukocytosis in PV patients [1].

• JAK2^exon12 mutations are specific to PV (not found in other MPNs), are usually associated with higher hemoglobin levels than in JAK2^V617F mutated PV, and typically present with isolated or “pure” erythrocytosis (without thrombocytosis or leukocytosis) [7, 8].

The mainstays of treatment include low-dose aspirin and maintenance of a hematocrit below 45%, either through phlebotomy or cytoreductive therapies (as we call the drugs used to treat MPNs by reducing blood cell production) such as hydroxyurea or interferon-alpha. JAK2 inhibitors, which primarily target dysregulated JAK-STAT signaling, are also used in selected patients. While hematocrit control reduces thrombosis risk [9], no clear benefit has been demonstrated regarding symptom burden or disease progression [10].

Other causes of erythrocytosis

PV is a primary acquired cause of erythrocytosis. Other causes of erythrocytosis can be classified as shown in Table 1.

Table 1.

Summary of the main causes of erythrocytosis.

	Congenital erythrocytosis (rare)	Acquired erythrocytosis
Primary	They are almost exclusively caused by mutations in the erythropoietin receptor (EPO-R) [124–126]. Some germline JAK2 mutations have also been described [127, 128].	PV, almost exclusively caused by JAK2 mutations. Other mutations, such as SH2B3 (LNK), have rarely been reported [129].
Secondary	They may result from hyperaffine hemoglobin variants, abnormalities in the hypoxia pathway (such as a gain-of-function mutation in the HIF-2α gene [130] or Chuvash erythrocytosis, caused by mutations in the VHL gene leading to constitutive activation of the HIF pathway [131, 132]), exceptional 2,3-diphosphoglycerate (2,3-DPG) deficiencies [133, 134], or PIEZO1 mutations [135]. A single-nucleotide deletion in EPO, which triggers overproduction of EPO via an alternative intronic promoter, has been reported [136]. More recently, it has been demonstrated that some cases can be related to variants in EPO leading to the production of hepatic-like EPO, with an atypical glycosylation pattern associated with increased EPO activity [137].	Associated with high serum EPO levels, they can be due to: − Central hypoxia (e.g., respiratory failure, sleep apnea, congenital heart disease with right-to-left shunt; or hypoxia related to living at high altitude), − Local hypoxia (e.g., renal artery stenosis), − Carbon monoxide intoxication secondary to water-pipe smoking, − Hyperproduction of EPO or EPO-like substances (e.g., renal or hepatic tumors), − Exogenous administration of EPO or high-dose androgens, or other treatments such as SGLT-2 inhibitors for diabetes.

VHL von Hippel–Lindau tumor suppressor.

PIEZO1 piezo-type mechanosensitive ion channel component 1.

The entity called idiopathic erythrocytosis (IE) is an absolute erythrocytosis of unknown cause, diagnosed by exclusion of primary and secondary erythrocytosis. Little is known regarding the clinical characteristics, natural history, and optimal management of patients with IE [11].

In a recent single-center study, patients with JAK2-unmutated unexplained erythrocytosis underwent next-generation sequencing using a hereditary erythrocytosis panel. Genetic variants were identified in 61% of cases, most commonly in PIEZO1 (20%), HIF1A (18%), and ANKRD26 (9%), the majority being variants of uncertain significance. No significant association was found between variant status and symptoms, erythropoietin levels, or thrombotic events [12].

Iron homeostasis and hepcidin

Hepcidin, encoded by the HAMP gene, is a 25-amino-acid peptide hormone produced predominantly by hepatocytes and is the master regulator of systemic iron homeostasis [13]. Hepcidin binds to and functionally inhibits or degrades ferroportin, the sole iron exporter. As a result, hepcidin reduces dietary iron absorption in the duodenum and iron release from macrophages and hepatocytes, limiting circulating iron availability [14].

Ferroportin (encoded by the SLC40A1 gene) is expressed on the surface of cells that are involved in iron absorption, recycling, and storage, especially on the basolateral surface of enterocytes and on macrophages. Its expression on macrophages is transcriptionally regulated by heme [15], iron [16], and inflammation [17]. Interestingly, ferroportin is also expressed on erythroid progenitors [18, 19]. This erythroid ferroportin transcript lacks an iron-responsive element (IRE) [20] and is not repressed in iron-deficient conditions. Increased hepcidin expression, by blocking ferroportin (therefore iron export) and increasing intracellular iron levels in RBC precursors, likely enhances RBC production and heme synthesis [18].

Hepcidin expression is regulated by several physiological signals. It is upregulated by increased plasma iron [21] and inflammation [22] – particularly interleukin-6 (IL-6) via the STAT3 pathway – and downregulated by iron deficiency, anemia, and erythropoietic activity [23].

Oxygen-sensing pathways also contribute to the regulation of iron and erythropoiesis through hypoxia-inducible factors (HIFs) [24]. Under normoxic conditions, HIF-α subunits are hydroxylated by prolyl hydroxylase domain (PHD) enzymes, particularly PHD2, encoded by EGLN1, an iron-dependent enzyme that targets HIF for degradation. In conditions of iron deficiency, reduced PHD2 activity leads to stabilization of HIFs, increased EPO production, and enhanced erythropoiesis. EPO is predominantly regulated by HIF-2α rather than HIF-1α, particularly in renal interstitial cells (responsible for hypoxia-induced EPO production) [25]. Notably, PHD2 is more specific for HIF-1, whereas PHD1 is more specific for HIF-2.

HIF represses hepcidin gene expression [26], by EPO-dependent or independent pathways [27–31]. The erythropoiesis-dependent suppression of hepcidin is primarily mediated by erythroferrone (ERFE), a hormone secreted by erythroblasts in response to EPO stimulation [32].

HIF activation has also been implicated in pulmonary hypertension and in the thrombotic complications observed in PV and Chuvash erythrocytosis [33, 34], and small-molecule HIF-2α inhibition reverses polycythemia and pulmonary hypertension in murine models [35].

Hepcidin regulation and action are summarized in Fig. 1.

Fig. 1.

Regulation and action of hepcidin.

The other actors of hepcidin regulation include the bone morphogenetic protein (BMP) signaling pathway, a key regulator of the iron-dependent regulation of hepcidin expression [36, 37]. The ligands BMP6 and BMP2, and the co-receptor hemojuvelin (HJV), stimulate BMP type I receptor (ALK 2/3) in iron overload states [37]. HJV, a membrane protein predominantly expressed in hepatocytes, functions as an essential BMP co-receptor that enhances BMP/SMAD1/5/8 signaling and promotes hepcidin transcription [38]. Loss-of-function mutations in HJV result in markedly reduced hepcidin levels and severe iron overload, similar to Hamp-related juvenile hemochromatosis [38]. The complex formed by transferrin receptor 1 (TfR1) and the High Fe²⁺ protein (HFE), together with transferrin receptor 2 (TfR2), can associate with the BMP receptor complex to modulate the BMP/SMAD1/5/8 signaling pathway and hepcidin transcription [37, 39] (illustrated in Fig. 2).

Fig. 2. Signaling pathways regulating hepcidin transcription in hepatocytes.

MT2: matriptase 2. Hjv: hemojuvelin. BMP: bone morphogenic protein. Hfe: human homeostatic iron regulator. Tf: transferrin. TfR1 and 2: transferrin receptor 1 and 2. EPO: erythropoietin. ERFE: erythroferrone. SMAD: suppressor of mothers against decapentaplegic. IL-6: interleukin 6. JAK: Janus kinase. STAT: signal transducers and activators of transcription. Hamp1: hepcidin antimicrobial peptide.

Notably, in addition to its role in hepatocytes as a regulator of hepcidin expression and iron homeostasis, TfR2 is also involved in the regulation of erythropoiesis as a partner of EPO-R [40]. In erythroid progenitors, TfR2 associates with EPO-R to modulate EPO sensitivity and proliferation according to iron availability. In murine models, loss of erythroid TfR2 enhances erythropoietic output and EPO responsiveness, further linking iron status and RBC production [41, 42].

Finally, matriptase-2 (encoded by the TMPRSS6 gene), a transmembrane serine protease, acts as a negative regulator of the BMP-SMAD pathway by cleaving HJV, thereby reducing BMP signaling and suppressing hepcidin expression [43].

Disruption of hepcidin regulation leads to iron disorders:

• Low hepcidin levels are responsible for the iron overload seen in HH (see below) and in ineffective erythropoiesis, where increased erythropoietic activity suppresses hepcidin production through ERFE production [44].

• High hepcidin levels can cause anemia by limiting iron availability to developing erythroblasts, as seen during chronic inflammation and iron-refractory iron deficiency anemia (IRIDA, an autosomal recessive disorder caused by mutations in the TMPRSS6 gene) [45].

Hereditary hemochromatosis (HH)

HH is the most common genetic iron overload disorder, predominantly caused by mutations in the HFE gene, particularly homozygous C282Y mutations [46]. HFE encodes a transmembrane protein that senses circulating transferrin-bound iron and regulates hepcidin expression in hepatocytes, thereby controlling intestinal iron absorption and iron release from macrophages. While C282Y mutations disrupt this function and cause the classical iron overload phenotype, the pathogenicity of the H63D variant is much less clear, with a milder or uncertain effect on iron metabolism [47]. At least five different categories of HH are known; the other (very rare) forms result from mutations in TfR2, HJV, HAMP, or ferroportin.

These mutations lead to inappropriately low hepcidin production. This insufficient production of hepcidin fails to downregulate ferroportin, leading to unregulated intestinal iron absorption. Iron efflux into the bloodstream thus continues despite elevated iron stores [14], resulting in increased transferrin saturation and non-transferrin-bound iron (NTBI) levels, and progressive accumulation of iron in organs such as the liver, pancreas, heart, and joints [48].

Clinically, HH may remain asymptomatic for years. However, chronic iron overload can eventually cause organ dysfunction: liver fibrosis or cirrhosis, diabetes mellitus, arthropathy, cardiomyopathy, and hypogonadotropic hypogonadism. The clinical penetrance is variable and often incomplete, particularly in women [49].

Standard treatment consists of phlebotomies based on iron overload (evaluated by liver MRI and plasmatic ferritin levels).

Iron homeostasis and erythrocytosis

Iron dysregulation in PV

Iron deficiency is very common in patients with PV [50], and can even hide the erythrocytosis – a condition known as masked PV [51–53]. The blood test reveals low ferritin levels, low transferrin saturation, and decreased mean corpuscular volume (MCV). Clinically, it may result in burdensome symptoms such as fatigue, impaired physical performance, cognitive impairment, glossitis, cheilosis, koilonychias, or restless legs syndrome, even in the absence of anemia [54]. These symptoms are reversible with iron supplementation [55] (although it is generally not recommended in PV as it promotes erythropoiesis, as discussed in II. 2.).

Noteworthy, iron deficiency is often exacerbated by therapeutic phlebotomies, whose aim is to restrict iron availability and thereby reduce erythropoiesis by limiting hemoglobin synthesis. Iron restriction also lowers the MCV, which contributes to reducing hematocrit levels.

The iron deficiency is frequently observed in PV patients before any phlebotomy, and can possibly be due to several coexisting factors (Fig. 3):

Fig. 3. Iron deficiency of any cause (left), and iron deficiency in Polycythemia Vera (right).

(* represents the starting event).

Increased need for erythropoiesis and aberrant iron restriction response

The persistence of erythrocytosis despite iron deficiency in PV can be counterintuitive. While iron deficiency is a logical consequence of increased erythroid proliferation, it would be expected to limit erythropoiesis. In iron-deficient states, several iron-sensitive pathways inhibit erythroblast proliferation and differentiation. These include mechanisms involving TfR2 [56, 57], EPO-R signaling [58], or erythroblasts’ cell cycle [59]. TfR2 acts as an iron sensor in erythroid progenitors, modulating EPO-R signaling according to iron availability. Under iron-deficient conditions, committed erythroid precursors slow their cell cycle and lose responsiveness to EPO [60], limiting erythroid expansion [56, 58].

In PV, however, clonal erythropoiesis appears partially resistant to these inhibitory signals as it persists despite iron deficiency. Clonal erythropoiesis keeps capturing systemic iron resources for hemoglobin synthesis, ultimately depleting iron stores.

Gastrointestinal bleeding

Chronic occult gastrointestinal bleeding is a frequent finding in PV patients [61, 62] and may be underdiagnosed, contributing to iron loss. Investigating and treating digestive bleeding may therefore be important in PV patients with unexplained iron deficiency.

Inflammation-driven hepcidin induction and impaired iron absorption

Although very few studies have directly assessed intestinal iron absorption in PV, this question has long been raised [63]. Iron absorption depends on its uptake by enterocytes via the apical iron transporter DMT1 (divalent metal transporter 1) and its export into the plasma, through the iron exporter ferroportin, which is inhibited by hepcidin.

In PV, hepcidin is expected to be low due to iron deficiency and increased erythropoiesis, theoretically promoting iron absorption. However, PV is often associated with chronic low-grade inflammation [64], which prevents a sufficient decrease of hepcidin [65]. IL-6, upregulated via JAK/STAT signaling, is a strong inducer of hepcidin expression [66] and is elevated in JAK2^V617F murine models. Treatment with JAK2 inhibitors normalizes IL-6 production [66].

By inducing hepcidin and promoting ferroportin degradation, inflammation restricts iron export from enterocytes and macrophages, thereby limiting dietary iron absorption and recycling. This inflammation may also explain why ferritin levels are sometimes higher than anticipated in iron-deficient PV patients, as ferritin expression is itself upregulated by inflammatory signaling.

Additional factors, possibly including impaired hypoxic responses (discussed in [67]), may also contribute to impaired iron absorption but remain incompletely understood.

Insufficient ERFE levels

While ERFE levels are increased in PV, they remain lower than in other conditions like β-thalassemia, where expanded but ineffective erythropoiesis leads to massive ERFE production and near-complete hepcidin suppression [32]. In PV, the low EPO levels and lack of accumulation of immature erythroblasts might limit ERFE output, making it unable to suppress hepcidin enough to correct iron deficiency.

Accordingly, in a PV mouse model, deletion of ERFE had little effect on hepcidin levels nor PV phenotype, demonstrating that ERFE is not a major hepcidin regulator in PV [68].

Moreover, studies in PV patients and mice with JAK2^V617F and JAK2^exon12 mutations suggest that different JAK2 mutations may influence iron metabolism differently [67].

In mice, JAK2^exon12 mutations (found mostly in pure erythrocytosis) are associated with higher ERFE levels and more pronounced hepcidin suppression than the JAK2^V617F mutation, despite similar iron stores [69]. JAK2^V617F and JAK2^exon12 PV mice are iron-deficient, but JAK2^exon12 mutants retain more iron, consistent with their stronger erythrocytosis, whereas ET-like JAK2^V617F mice maintain normal iron levels. Iron availability modulates the ET–PV phenotype by shaping megakaryocyte–erythroid progenitors (MEP) fate: low iron suppresses erythropoiesis and shifts MEPs toward platelet production, while high iron enhances erythropoiesis and exacerbates PV features [70].

In PV patients, those with a JAK2^exon12 mutation are more likely to have iron deficiency than those with the JAK2^V617F mutation. One hypothesis would be that they consume iron faster since they present higher levels of erythrocytosis. Among patients with JAK2^V617F mutation, a higher allele burden (variant allele frequency (VAF) ≥ 50%) was associated with a higher probability of iron deficiency [50].

Treatment of iron deficiency in PV

The management of iron deficiency in PV is complex. On one hand, iron supplementation could relieve the often-distressing symptoms commonly associated with iron deficiency, providing meaningful improvement in patients’ quality of life. On the other hand, iron depletion is precisely the therapeutic goal of phlebotomy, and iron supplementation, by stimulating erythropoiesis, would therefore be counterproductive.

If a cause of iron deficiency, such as chronic bleeding, is found, etiological treatment is, of course, recommended.

Cytoreductive therapy, recommended for high-risk PV patients, may be associated with improvement of iron deficiency. It often improves symptoms more significantly in iron-deficient than in iron-replete individuals [67]. Some therapeutic strategies in PV, particularly the use of JAK2 inhibitors [71], may alleviate PV-related symptoms partly by restoring iron availability. However, iron deficiency and PV-related symptoms can overlap, which makes this assessment challenging.

Iron overload, HH and erythrocytosis

Hematological parameters in HH patients

Iron overload has been associated with various hematological perturbations, ranging from abnormalities in erythrocytes, leukocytes, or platelet functions [72] to malignant hemopathies [73].

The increased iron bioavailability may enhance erythropoiesis by increasing iron uptake by erythroid precursors, thereby stimulating RBC production. A correlation between erythrocytosis and HH has been suspected for a long time [74]. Some studies have indeed reported that patients with HFE mutations, even heterozygous, have significantly increased RBC, hemoglobin concentration, hematocrit, and MCV compared to controls, even though overt erythrocytosis is rare [75–79]. Other studies reported normal erythroid parameters in HH, yet observed that anemia seems very uncommon in these patients [80]. Iron appears to support erythropoiesis, but it remains unclear whether it could intrinsically drive it beyond physiological limits in the absence of erythropoietic stimuli, maybe through a potential EPO-independent mechanism mediated by intracellular iron signaling and heme synthesis.

On the other hand, it is known that iron overload induces oxidative stress [81]. Oxidative stress, in turn, can be toxic to erythropoiesis [82]. One could thus suggest that under conditions of iron overload, erythropoiesis may be slightly suppressed due to oxidative stress from iron toxicity, potentially counterbalancing iron’s stimulatory effects on erythropoiesis.

Association between HH and PV

HFE mutations in PV patients

In PV, previous studies have not consistently reported an increased prevalence of HH mutations [83, 84], although case reports indicate that the coexistence of PV and HH is possible [85–87].

In a series of 326 chronic MPNs including 175 PV patients, HFE C282Y mutations were found less frequently in patients than in controls (996 blood donors), suggesting, according to the authors, a potential protective effect of these mutations against MPNs [88]. However, this study included a heterogeneous MPN population and had a relatively small sample size, which restricts the interpretation of its findings specifically for PV.

By contrast, a genome-wide association study conducted in two very large cohorts of 440 and 394 PV patients, compared with 403,351 and 217,902 controls, identified single-nucleotide polymorphisms in the HFE gene (particularly the C282Y mutation) as strongly associated with PV [68]. Importantly, no association was found with other MPNs (essential thrombocythemia and primary myelofibrosis). The authors suggest that HFE mutations may increase iron availability by lowering hepcidin levels, thereby facilitating uncontrolled erythropoiesis once somatic JAK2 mutations occur [68, 89].

However, most studies report associations, and a causal relationship between HH and PV has not been formally established in patients.

JAK2 mutations in HH patients

There is no evidence that patients with HH develop PV or other MPNs at higher frequency. In a small series of 48 HH patients, no JAK2^V617F mutation was detected [90].

Role of phlebotomies for HH in the emergence of PV

Interestingly, the case of a patient developing PV after phlebotomy treatment for HH has been reported; however the JAK2^V617F mutation was already present before phlebotomies [91].

In a series of PV patients, the prevalence of former blood donors was about three times higher than in the general population. Blood donation could simply delay PV diagnosis, but the authors also suggest that regular blood loss may stimulate erythropoiesis [92]. Nonetheless, there is no evidence that phlebotomies performed for HH patients constitute a risk factor for developing PV.

Association between iron overload and erythrocytosis

HFE mutations in erythrocytosis

In idiopathic erythrocytosis (IE), a significant number of patients have high or borderline-high ferritin levels [93], suggesting an association between iron balance and erythrocytosis. Moreover, a relevant number of IE patients carry HFE mutations [94–96]: up to 44% [93] or even 55% of patients [97], depending on the study, even though these mutations can be heterozygous and are not always associated with an increased risk of HH. Independent of HH, certain HFE variants can have an impact on erythropoiesis by altering iron content and EPO function [98].

Notably, HFE-knockout mice exhibit increased hematocrit and hemoglobin levels compared with HFE wild-type mice, suggesting that HFE mutations may stimulate, or at least facilitate, RBC production [99].

HFE mutations have also been reported at a markedly higher frequency in elite French athletes (80%) compared with the general population (27%), suggesting that increased iron availability may enhance endurance performance by promoting erythropoiesis [100].

Iron homeostasis variants and idiopathic erythrocytosis

Iron overload, whether due to HH or other causes, may represent a cause of unexplained erythrocytosis. In this line, our team has indeed observed that mouse models of iron overload (either through hepcidin knockout or oral iron gavage) develop significant erythrocytosis (unpublished data).

Further studies are needed to better define the threshold at which iron begins to stimulate erythropoiesis and to clarify the underlying mechanisms.

Among possible mechanisms, in addition to HFE, variants in other iron metabolism genes may represent risk factors for developing erythrocytosis. Interestingly, TfR2 variants have already been identified in IE patients [101]. Functional studies support a tissue-specific model: in hepatocytes, stabilized TfR2 in iron-replete states promotes hepcidin expression, whereas TfR2 destabilization in iron deficiency reduces hepcidin signaling. In erythroid progenitors, TfR2 limits EPO sensitivity when iron is sufficient, while its destabilization under iron restriction enhances EPO responsiveness and erythroid proliferation. These complementary roles indicate that TfR2 variants could influence RBC production depending on iron status [41, 42].

In PV, despite systemic iron deficiency, expanded erythroid precursors overexpress TfR1, thereby facilitating maximal iron uptake to support hemoglobin synthesis [67, 102]. To our knowledge, no TfR1 variants have been described in either PV or IE. The only reported mutation within the endocytosis motif of TfR1 in humans has minimal effect on erythropoiesis but rather results in immunodeficiency [103].

Another protein of interest is CD44, a cell adhesion molecule highly expressed on early hematopoietic progenitors and certain megakaryocyte–erythroid progenitor (MEP) subpopulations. CD44 is upregulated in stress erythropoiesis and may support excessive erythroid differentiation in PV [102, 104]. Interestingly, it has been recently found that CD44 may contribute to iron uptake by tumorigenic cells through the endocytosis of iron-hyaluronate complexes [105]. CD44 is expressed at an earlier stage of erythroid progenitor development than TfR1 [102, 106, 107]. Its expression is increased by iron, unlike TfR1; CD44 lacks an iron-responsive element (IRE) and is therefore not subject to negative feedback regulation [105]. One could hypothesize that, under conditions of iron overload, CD44-mediated iron uptake may lead to a “pre-loading” of erythroid progenitors, potentially stimulating erythropoiesis. It would be interesting to investigate whether variants in CD44 could represent an unrecognized cause of erythrocytosis (Fig. 4).

Fig. 4.

Possible role of iron overload, or increased iron uptake, on erythropoiesis.

New therapeutic approaches targeting hepcidin in PV

Hepcidin in PV

Hepcidin is controlled by several signals, including circulating iron levels, ERFE (secreted by erythroblasts), and inflammatory cytokines [20]. Hepcidin acts by limiting iron efflux into plasma by blocking ferroportin.

As discussed before, in PV, hepcidin levels are supposed to be low due to sustained erythropoiesis, persistent JAK/STAT signaling, and elevated ERFE, which should promote iron mobilization. However, PV is also associated with low-grade chronic inflammation, which can stimulate hepcidin, leading to a higher-than-expected hepcidin level (abnormally normal), insufficient to correct the functional iron deficiency [67].

Regarding therapy, phlebotomies worsen iron deficiency, and classical cytoreductive treatments (hydroxyurea, interferon-alpha) have no known effects on iron homeostasis.

Ruxolitinib, a JAK1/2 inhibitor, has been shown to improve iron parameters and hepcidin levels in PV patients with baseline iron deficiency [71], as compared to those receiving hydroxyurea or interferon-alpha. This may result from an action of ruxolitinib on the chronic low-grade inflammation associated with PV through JAK1 inhibition.

Momelotinib, another JAK1/2 inhibitor inhibitor which was initially reported as an ACVR1 inhibitor (BMP receptor, type 1), has been shown to improve anemia in patients with primary myelofibrosis. In a rat model of anemia of chronic disease, this effect was driven by direct inhibition of ACVR1-mediated hepcidin expression in the liver, leading to increased mobilization of sequestered iron from cellular stores and subsequent stimulation of erythropoiesis [108]. However, this effect remains controversial, and it is also possible that momelotinib exerts a weaker inhibition of JAK2. In clinical studies, momelotinib has been associated with a more modest rise in ferritin levels over time relative to ruxolitinib, a pattern that may indicate distinct effects on iron handling and storage.

Increasingly, treatments targeting hepcidin more directly are being studied in PV.

Increasing hepcidin in PV

Increasing hepcidin in PV could limit iron availability for erythropoiesis (potentially reducing the need for phlebotomies and cytoreductive therapies), while preserving iron stores (by sequestering it in hepatocytes and macrophages), thereby improving quality of life by alleviating systemic iron deficiency symptoms [109].

Preclinical studies have demonstrated that administration of exogenous hepcidin in JAK2^V617F murine models effectively reverses erythrocytosis and reduces splenomegaly [110].

Similar results have been obtained with another hepcidin mimetic agent: an antisense oligonucleotide against TMPRSS6 mRNA (a negative regulator of hepcidin production), which leads to increased endogenous hepatic hepcidin expression. In JAK2^V617F mice, this treatment resulted in decreased RBC counts and hematocrit levels, and reduced erythroblasts in the bone marrow [111]. Other treatments have been used with the same success, such as a parenteral synthetic hepcidin [112] and an orally bioavailable ferroportin inhibitor (vamifeport) [70]. This latter option is of specific interest as it has been shown that erythroid progenitors express a ferroportin transcript, which lacks an iron-responsive element (see I.3.) [18].

Building on these findings, several therapeutic approaches are being explored to either mimic hepcidin function or enhance its endogenous production. These include full-length synthetic hepcidin peptides, truncated analogues known as mini-hepcidins, stimulators of endogenous hepcidin expression (such as BMP pathway agonists), and ERFE antagonists [113].

Rusfertide, a hepcidin mimetic, has been evaluated in PV patients and is a very promising approach to replace phlebotomies in low-risk PV patients, possibly delaying the need for cytoreductive therapy [114–118]. The phase 3 VERIFY trial of rusfertide (ClinicalTrials.gov number, NCT05210790) is ongoing.

Several hepcidin agonists (rusfertide, but also divesiran - a liver-targeted GalNAc-conjugated double-stranded 19-mer siRNA against TMPRSS6, and sapablursen - an antisense oligonucleotide targeting TMPRSS6) are being developed to control erythrocytosis in PV [119].

A humanized monoclonal antibody against TMPRSS6, DISC-3405, is also under development. In preclinical models and in healthy volunteers, DISC-3405 has been shown to stimulate endogenous hepcidin production, reduce serum iron, and lower hematologic parameters [120]. A phase II study is ongoing in PV.

By selectively restricting iron availability to the erythroid compartment without inducing systemic iron deficiency, all these agents aim to replicate the physiological effects of hepcidin while avoiding the adverse consequences of repeated phlebotomies [109].

Hepcidin antagonists in PV?

While hepcidin mimetics are being explored to limit erythropoiesis in PV, the opposite approach – using hepcidin antagonists – may alleviate iron deficiency symptoms.

Indeed, patients with PV often develop functional iron deficiency, which can lead to burdensome symptoms [54]. Elevated or inappropriately normal hepcidin levels may contribute to limited iron availability.

Hepcidin antagonists, including neutralizing antibodies, anticalins, anti-hemojuvelin monoclonal antibodies (DISC-0974), or small molecules that inhibit hepcidin production or block its interaction with ferroportin, are under investigation in situations associated with high hepcidin levels, such as anemia of chronic disease and iron-refractory iron deficiency anemia (IRIDA) [121, 122]. These agents may also be beneficial in anemic patients with primary myelofibrosis, a disease characterized by increased hepcidin levels [119], including those receiving JAK1/2 inhibitors. Early clinical trials with anti-hemojuvelin monoclonal antibodies are undergoing [123].

In contrast, the use of hepcidin antagonists in PV requires caution, as increasing iron availability may enhance erythropoiesis in a disease already driven by hyperactive erythroid proliferation. Such interventions should therefore be limited to selected cases of severe symptomatic iron deficiency or combined with cytoreductive therapy to mitigate the erythropoietic response.

Interestingly, ligand traps (such as luspatercept, elritercept) of the TGF-β family used in myelodysplastic syndromes are now used in certain MPNs. These ligand traps enhance erythropoiesis by neutralizing specific TGF-β superfamily ligands (mainly activins and GDF11) that inhibit late-stage erythroid maturation. By reducing SMAD2/3 signaling in erythroid precursors, they promote effective RBC production. This mode of action differs fundamentally from regulators of hepcidin, whose activity depends on the BMP–SMAD1/5/8 pathway, driven chiefly by BMP6 and BMP2 to control iron metabolism and ferroportin activity. Thus, TGF-β ligand traps act on erythroid maturation rather than on hepcidin-mediated iron regulation, engaging distinct branches of the TGF-β signaling network.

Conclusion

Iron metabolism is central to the pathophysiology and management of erythrocytosis, especially PV.

Functional iron deficiency is common in PV, driven by repeated phlebotomies, chronic inflammation, and inadequate hepcidin levels. Despite intense erythroid proliferation, PV patients often present with low ferritin and higher-than-expected hepcidin levels, reflecting a maladaptive iron-restricted state. This not only influences erythropoiesis but also contributes to symptoms altering the patients’ quality of life.

New therapies that modify systemic iron regulation by targeting hepcidin are emerging. Hepcidin mimetics may help control erythrocytosis and reduce phlebotomy needs, while antagonists could relieve symptomatic iron deficiency in selected cases. Personalized approaches targeting iron metabolism may complement or even reduce the need for cytoreductive treatments, though long-term safety and efficacy remain to be established.

On the other hand, iron overload could act as a trigger for erythroblasts’ proliferation and be involved in cases of idiopathic erythrocytosis. The association between hemochromatosis and erythrocytosis remains an intriguing area of research.

Author contributions

Conceptualization: GF, Literature search: BG, CP, IP, SV, GF, Writing, original draft: BG, GF, Writing, review & editing: BG, CP, IP, SV, GF, Visualization: BG, GF.

Funding

The authors declare that they received no funding in relation to this manuscript.

Competing interests

The authors declare no competing interests.