The molecular pathogenesis of BCR::ABL-negative myeloproliferative neoplasms was unascertained until 2005, when the presence of an activating point mutation in the JAK2 gene was first described [12,13,14,15]. Thereafter, other driver mutations events have been elucidated, such as the presence of mutations in MPL (which encodes the thrombopoietin receptor, TPOR) [16] or CALR [17,18]. These three genes are considered disease drivers as, when mutated, they could provoke MPN development in murine models [19].

The canonical and most common JAK2 mutation is a single nucleotide change at codon 617, located in exon 14, resulting in the substitution of a valine for phenylalanine (V617F). This amino acid shift in the JH2 pseudokinase domain prevents the autoregulatory inhibition of JH1, resulting in the constitutive phosphorylation and activation of the Janus kinase [13]. Roughly one third of PV patients who are JAK2V617F-negative harbor typically insertions or deletions in JAK2 exon 12 [20]. Both mutations lead to a gain-of-function of the kinase, the consequent activation of its downstream targets and deregulation of JAK/STAT signaling [5]. Moreover, the JAK2V617F allele burden has been associated with the MPN subtype, progression to secondary MF and an increased thrombotic risk [6,21,22].

The conventional MPL mutations take place in W515, located in exon 10, resulting in amino acid substitutions [23], with W515L/K being the most common changes [24]. Serine 505 could be also mutated, though less frequently [25]. All these mutations prevent the inactive conformation of the TPOR receptor, leading to its cytokine-independent activation and the constitutive JAK/STAT pathway activation [26].

CALR mutations present in MPN patients are insertions and deletions located at the exon 9, generating a novel C-terminal peptide sequence [17]. Almost 85% of CALR-mutated patients harbor a deletion of 52 bp (type I, CALR^del52) or an insertion of 5 bp (type II, CALR^ins5) [26] and drive JAK/STAT pathway activation through TPOR activation [11].

Although driver mutations were classically considered as mutually exclusive, some studies have identified double mutated patients (i.e., JAK2V617F and CALR, JAK2V617F and MPL, CALR and MPL, and JAK2V617F and JAK2 exon 12 mutations) [27], arising from different HSC clones. Recent work suggests that CALR mutations tend to occur later in life than JAK2V617F, which would explain the higher proliferative advantage of the CALR malignant clone compared to JAK2V617F [28].

Finally, in recent years, there has been a growing body of evidence suggesting, in some cases, an in utero or postnatal acquisition of MPN-driver mutations [29]. In line with this, it has been estimated that the JAK2V617F mutation could occur in a single HSC several decades before MPN diagnosis [30,31]. A better understanding of the evolutionary dynamics of MPNs, in addition to the long disease latency, might open up opportunities for early intervention.

Considering that these three mutations have remained the main drivers of the disease for years, patients who exhibit histological and phenotypic MPN features but lack either of these mutations have classically been known as “triple negative” (TN) patients. Around 10–15% of ET and PMF patients are triple negative [32]. The investigation of these TN patients with next generation sequencing (NGS) approaches showed that many other mutations may affect MPN patients, especially those classified as TN [5,11]. Furthermore, delving deeper into this particular group of patients, sequencing studies have revealed that TN ET patients showed the dysregulation of genes involved in MAPK, tumor necrosis factor and NF-κB pathways, leading to activation of the JAK/STAT pathway [33].

Some relevant patient cohort studies have sequenced 104 genes in 197 MPN patients and followed clonal evolution relating the number of somatic mutations identified with reduced overall survival and an increased risk of AML transformation [34]. Larger studies sequenced coding exons from 69 myeloid cancer genes in 1887 patients [6] and developed a new personalized prognosis classification for MPNs. Forty-five percent of patients from this cohort showed mutations in JAK2, CALR or MPL, whereas the remaining 55% were classified as TN MPNs. Authors defined eight genomic subgroups with different clinical phenotypes creating a prognostic model tool.

In general, besides JAK2, CALR or MPL mutations, many others contribute to the MPN phenotype. As previously mentioned, the presence of additional mutations is usually more common in advanced stages of the disease (MF) than in earlier chronic stages (PV and ET) [5,6]. The most affected genes are involved in DNA methylation, mRNA splicing and signaling with mutations in TET2, DNMT3A and ASXL1 found in almost half of patients [1] but also EZH2 [35], CBL, PPM1D, SF3B1, NFE2, TP53, SRSF2 and U2AF1 [6] as the most frequent ones.

Mutations in DNMT3A, TET2 and ASXL1 (DTA) genes confer a selective advantage that may lead to an expanded clone of cells, termed clonal hematopoiesis, which is an age-related pre-malignant condition. The mutations associated with clonal hematopoiesis of indeterminate potential (CHIP) increase cardiovascular risk and associated mortality [36,37]. The study of additional CHIP mutations and their association with increased cardiovascular risk has attracted a great deal of research interest in MPNs, and some authors suggest that DTA mutations, particularly TET2 mutations, may be an independent risk factor for thrombosis in PV [38]. In contrast, the presence of mutations in ASXL1, RUNX1 or EZH2 seems to play a salutary effect on the risk of arterial thrombosis, particularly in ET patients [39].

Although in previous studies on MPNs, mutations in genes involved in DNA methylation had not been related to blast phase progression [40], it is of relevance that they have been described as a risk factor for relapse after treatment for AML [41]. Indeed, more recent studies in a large cohort of Japanese PV and ET patients identified ASXL1 mutations as a risk factor for leukemic/myelofibrotic transformation [42].

Furthermore, ASXL1 [43] or EZH2 mutations [44] by themselves have been related to poorer survival in PMF patients. In fact, when both mutations coexist, MPN patients present a higher risk of secondary MF transformation and poorer survival [45]. Therefore, mutations in these genes have been included as high risk mutations (HRMs) in the Mutation-Enhanced International Prognostic Score System for Transplantation-Age Patients With Primary Myelofibrosis (MIPSS70 score) [46].

Despite being particularly more common in myelodysplastic syndrome (MDS) patients [47], mutations in the splicing factors SF3B1, SRSF2 and U2AF1, between others, represent the second most frequently mutated gene category in MPNs [48], especially in PMF, and have been linked to a higher risk of AML transformation as well as lower overall survival [45]. Notably, whereas SF3B1 is the most frequent mutated splicing factor in MPNs [49], different outcomes could be observed for each MPN entity. When mutated, it is associated with lower overall survival in ET [50] and secondary MF (particularly post-ET patients) [51] but not in PV [50] or PMF patients [52]. Specially, it is also related to lower myelofibrosis-free survival [50]. Conversely, SRSF2 mutations in all MPN patients predict adverse outcomes [50,53], and they are considered HRMs in the MIPSS70 score [46].

TP53 mutations are also particularly very common in MDS patients [54]. Although their presence in MPNs is lower, patients harboring TP53 mutations have a dismal prognosis and a higher risk of AML transformation [6,34,50], especially when accompanied by a proinflammatory microenvironment [55]. Not only TP53, but also mutations in components of other signaling pathways, such as PPM1D or NRAS, have been related to leukemic transformation and early death [35].

Altogether, in the last decade, genomic studies have identified multiple recurrent somatic mutations in MPNs. It is, therefore, intriguing how the same genetic background may have completely different disease courses, with completely different phenotypes, responses to treatment and risks of thrombotic events or leukemic progression. This fact underlines the importance of evaluating MPN biology at other levels beyond the mutational study in order to be able to anticipate the worsening of patients or administer the best therapeutic options in each case in a precise manner.

Although the elucidation of the molecular landscape of MPNs has been mainly focused on the detection of driver mutations, the presence of one of the three main driver mutations (JAK2, CALR and MPL) can lead to different phenotypes, demonstrating the influence of other factors in the development and progression of these neoplasms. Thus, the phenotype is not solely determined by the genomic profile, and disturbances in DNA transcription to mRNA or its translation to protein may be even more important.

One of the studies that best represents the important role of transcriptomic biology was carried out by Ding and colleagues who observed and evaluated the different treatment response of a monozygotic twin pair with childhood PMF [56]. The brothers shared the same genetic information, lacked the presence of mutations in JAK2, CALR or MPL but harbored mutations in SRSF2 and SF3B1, which could drive MF as well, as previously reported. Both patients received the same treatment, but unfortunately, only the younger brother responded. Notably, the mutational status did not change after treatment in either of them, suggesting that the differences in response should be explained by epigenomic or transcriptomic changes. In order to demonstrate this, the authors carried out a comparison of the transcriptomic profiles, which showed that the patient who responded presented with both the activation of the JAK/STAT signaling pathway and a clear influence of the cell cycle and apoptosis pathway through treatment. Particularly, after a transcription factor enrichment analysis, these authors showed that E3F and GATA were the transcription factors responsible for the differential expression of 40 genes after treatment, suggesting that their regulation could explain the differential treatment response between the brothers [56].

Indeed, other transcriptomic studies have also revealed the possible role of GATA1 in myeloproliferative neoplasms. One of them described the enrichment of two genes, GPR56 and RAB4a, together with the lower expression of TAL1 due to the increased phosphorylation of cytoplasmic kinases JAK/STAT, PI3K and GATA1 pathways in expanded erythroblasts from PV patients [57]. In the same line, a single cell RNA-seq study of 15 MF patients and 6 controls revealed differential expression patterns of the transcription factors, GATA1 and GATA2, among others, in the regulation of megakaryocyte–erythroid cell fate decisions [58].

Transcriptomic studies have also been carried out to evaluate the gene expression profile associated with driver mutations. For example, Berkofsy-Fessler and colleagues compared the transcriptional profile associated with the JAK2 canonical mutation [59]. The authors identified a panel of 14 JAK2-dependent genes, such as KLF4, which is implied in the pluripotency maintaining of embryonic stem cells and a panel of 12 JAK2-independent genes. These two panels allowed them to distinguish between PV and controls and between ET or PMF and the control group. Other studies have related transcriptomic aberrations of mesenchymal stromal cells (MSCs) from patients with JAK2V617F-mutated ET patients to WDR4 low expression with increased proliferation, reduced senescence and differentiation [60]. In contrast to both studies is a microarray gene expression analysis in 93 patients with MPNs (28 PV, 47 ET and 18 MF) in comparison to granulocytes of 11 age-matched normal donors; although this led to the identification of a characteristic gene expression profile related to JAK2 activation in MPN patients, it did not identify a specific expression signature derived from the JAK2-mutational state [61]. However, CALR-mutated patients presented a gene signature associated with activated JAK2 signaling (such as an upregulation in JAK2 and STAT1) [61]. This fact supports the efficacy of JAK-targeted therapies in MPNs regardless of genotype. From another perspective, Zini and colleagues compared the transcriptomic profile of CD34+ cells from ET patients harboring JAK2 or CALR mutations [82]. As expected, the authors described that different pathways were activated depending on the driver mutation (i.e., mTOR, MAPK/PI3K and MYC pathways), and despite of this, the same phenotype was shown, thus reinforcing the influence of transcriptomics on the disease.

In addition to exploring gene expression programs derived from the presence of known mutations, another main goal of wide-ranging studies based on transcriptomic data is to discover new targets or biomarkers of progression in the disease [62]. As an example of this, several studies have shown the dysregulation of inflammatory pathways in MPN patients. An analysis of PMF CD34+ and megakaryocytes (MK) cell transcriptomes showed deregulation of the MAPK pathway along with FLT3 expression [63]. A higher proportion of circulating FLT3+ CD34+ cells exhibited an increased MAPK effector phosphorylation independently of the JAK2V617F mutation in PMF patients. The activation of the FLT3 axis in PMF MK cell cultures induced the activation of the p38-MAPK cascade and overexpression of its targets (NFATC4, p53, AP-1 and IL-8), resulting in an inflammation context. As a consequence, reducing p38MAPK phosphorylation through FLT3 inhibition demonstrated the megakaryopoiesis process was improved [63].

Others have also identified a number of gene expression changes associated with inflammatory pathways and progression to fibrosis, distinguishing between pre-fibrotic MPNs and overtly fibrotic MPNs [64]. For example, the increased expression of several prefibrotic growth factors, matrix metalloproteinases, VEGFA, IGFBP7 and cell cycle regulators (CCND1, CCNA2, CCNB2 and CCNF) have defined a specific transcriptional signature in association with fibrosis development [65] and advanced stage disease [62]. In the same way, other authors also found the altered gene expression of cell cycle regulators CCND1, H2AFX and CEP55 as a fibrosis-signature in MF patients [66]. Similarly, the transcriptional alterations of BM-MSC in PMF patients revealed high TGFβ1 signaling, driving to osteogenic potential [67]. Importantly, the inhibition of the TGFβ1 receptor abrogated the osteogenic differentiation of MSC, suggesting its use in combination with the inhibition of hematopoietic cell proliferation as a novel therapeutic strategy in MF.

Apart from inflammatory or fibrotic pathways, other studies have found different novel targets. Using microarray analysis, the gene expression of CD34+ stem cells from the peripheral blood (PB) of idiopathic MF was compared to normal CD34+ controls [68]. The analyses revealed 174 differentially expressed genes, of which 8 (CD9, GAS2, DLK1, CDH1, WT1, NFE2, HMGA2 and CXCR4) were differently expressed in the two studied populations. Among them, WT1 expression was associated with a more active disease. Years later, Muggeo and colleagues interrogated the involvement of the pre-B-cell leukemia homeobox 1 (PBX1) in human MPNs [69]. Using a genetically modified knockout (KO) model, the authors found that when PBX1 is removed, the expression levels of genes upregulated in MPNs were downregulated in the murine model. Therefore, PBX1 might act downstream of JAK2, proposing a new target subjected to novel therapies.

Finally, based on different approaches, the evaluation of the mutational profile of 113 MPN patients using transcriptomic data have helped to predict around 149 novel neoantigens in 62% of the evaluated patients, which could be very useful as a source for generating a personalized vaccine or developing an adaptive cell therapy [49].

Other studies identifying new candidate prognostic markers or potential therapeutic targets have described the possible role of MAPK14 in PV [70], EPB42, CALR, SLC4A1 and MPL [71], LCN2, JAK2, MMP8, CAMP, DEFA4, LTF, MPO, HBD, STAT4, EBF1 [72] in PMF, CDH6, EHD2, FUT1, KIF26A, LINC00346, PTPRN, SERF1A, SLC6A9 [73], and CALR in ET [74].

The last part of this section includes transcriptomic studies that have evaluated alterations at the regulatory or processing level or RNA transcription. It is well known that microRNAs (miRNAs) regulate hematopoiesis, so their aberrant expression could prompt very diverse hematopoietic alterations. MiRNA deregulation in MPN patients has been known for years [83]. While some studies have described a 40-miRNA signature in platelets (PLT) of JAK2V617-negative [75], others have revealed the specific role of some miRNAs in MPNs. Thus, recent studies have revealed the role of miR-543 [76] and miR-382-5p [77] in MPNs. Furthermore, the deregulation of miR-146-a, a brake in NF-κB signaling, could drive a MF-like phenotype in a model of miR-146-a^KO mice [78].

Along the same lines, alterations in the 3′ untranslated regions (3′-UTRs), which are involved in mRNA stabilization and processing, may be implicated in the progression of many cancers such as MPNs [84]. Interestingly, new non-canonical CALR mutations in the 3′-UTRs, which are usually not detected by conventional techniques, have been described in JAK2V617/exon 12 mutation-negative MPN patients that resemble PV due to increased erythroid maturation [85]. By another way, as mutations in some splicing components are usual in myeloid malignancies and driver events in MPNs [6], some efforts have been focused on evaluating splicing aberrancies in splicing-mutated patients. In this sense, Schischlik and colleagues [49] demonstrated that MPN patients harboring SF3B1 mutations provided distinct 3′ splicing patterns with 250 new genomic alterations. However, even in the absence of splicing mutations, splicing anomalies could occur. A notable example is the detection of the JAK2 isoform lacking exon 14 in MPN patients [80,81].

Similar to transcriptomic studies, proteomic analyses have mainly focused on studying molecular changes resulting from the presence of certain mutations, the discovery of new targets and the pharmacological responses.

Since they were developed, proteomic profiling tools have helped to gain insights into the downstream molecular mechanisms activated by a particular mutation or pathway. The characterization of JAK2V617F signaling in MPNs is obviously of particular interest. The first study evaluating this was carried out by Mossuz and colleagues [95]. The authors investigated the influence of the JAK2 canonical mutation in ET patients using mass spectrometry-based analysis and found that the mutation did not significantly impact the serum proteome, suggesting that its presence partially influenced the ET phenotype and that other additional factors may be involved. However, the cell proteome characterization from wild-type and V617F-mutated JAK2 cells identified disruption changes in proteins from the TGFβ, p53 and MYC signaling as the most affected [86]. Later on, similar comparisons showed kinase motifs for glycogen synthase kinase-3 (GSK3), ERK and cyclin-dependent kinases (CDKs) enriched in JAK2-mutated cells, as well as increased phosphorylation of many proteins associated with the gene ontology terms, mRNA splicing and processing, as the main JAK2-mutated downstream signaling pathways [87].

In addition, protein interactome derived from CALR wild-type and mutated MPN cells were characterized, identifying higher CALR levels in the mutated cells and the increased recruitment of FLI1, ERP57 and CALR to the MPL promoter to enhance its transcription and, therefore, expression [88]. Similar approaches have helped to elucidate SCF (stem cell factor)/cKIT interplay in primary erythroid progenitors, using phosphoproteomic profiling of adult blood (AB), cord blood (CB) and polycythemia vera cells [89].

Proteomic analyses have also identified new targets as new therapeutic vulnerabilities in MPNs. In 2013, Gallardo and colleagues described the possible role of heat shock protein 70 (HSP70) as a key role in the proliferation and survival of the erythroid lineage in PV patients using two-dimensional gel electrophoresis (2D-DIGE) and mass spectrometry, suggesting its regulation as a potential therapeutic target [90]. In addition, new targets have been defined using murine models. In this sense, some researchers described the role of BM’s specific loss of ABI1 as a driver of MF disease, which was validated in MPN patient samples [91]. In this MF murine model, the phosphoproteomic studies showed that STAT3, ERK1/2 and AKT phosphorylation was significantly increased via the hyperactivity of SFK (family Src kinase)/STAT3/NF-kB signaling pathways. Particularly, Abi-1^KO mice significantly presented with an overexpression of genes involved in the NF-κB pathway. On the other hand, Jayavelu and colleagues described, in their MPN murine model, the loss of YBX1 as a sensitization mechanism for the inhibitory capacity of JAK inhibitors (JAKi), as genetic YBX1 disruption helps diminish JAK2i-persistent cells [87]. Additionally, as YBX1 could be dependently phosphorylated via JAK2V617F and MEK/ERK pathways, a combination therapy based on ruxolitinib and trametinib (a MEK inhibitor) was evaluated in PDX (patient-derived xenograft) mice models, proving that cell growth was decreased and molecular remission reached compared to ruxolitinib-treated PDX mice. In the same direction, other authors have also suggested that MAPK or cKIT treatment in PV could enhance JAKi responses based on their results [89]. In addition, the simultaneous treatment of MYC and p53, based on JQ1 and nutlin combination, in JAK2V617F cells showed the potential of this dual therapy in MPNs [86].

In addition, as previously noted [87], another advantage of proteomic studies is the evaluation of treatment responses. In this sense, a very recent study evaluated JAK inhibitor responses and described a central role for the activation of the MAPK survival pathways in persistent MPN cells, suggesting MEK/ERK pathway inhibition to get better treatment responses [92]. Particularly, Brusson and colleagues have previously reported that membranes of PV red blood cells presented an abnormal expression of several proteins, with some of them related to adhesion, such as Lu/BCAM (Lutheran/basal cell adhesion molecule) [93]. Later research from the same group [94] evaluated the effects of hydroxycarbamide (HC, hydroxyurea) and interferon-α (IFN) treatment in MPN patients applying proteomic studies. The authors described that both treatments diminished CALR levels, but neither of them up to the control levels. However, HC treatment enhanced the expression of proteins related to erythrocytes adhesion, such as Lu/BCAM and CD147, whereas IFN did not. Therefore, the results suggested that this altered expression maybe prompted a negative impact on vascular risk event occurrence, and HC treatment had adverse effects on red blood cell physiology. However, further studies assessing the actual risk of HC treatment in PV are needed.

Similarly to the intrinsic mechanisms involvement in the disease, extrinsic mechanisms may provide the signals that are necessary for the survival of MPN cells [10,96,97,98,99]. These mechanisms are also essential for MPN development, evolution or treatment resistance.

Since the first description of the relationship between Epstein–Barr virus and Burkitt’s lymphoma [100], other microorganisms infecting the human body have been linked to cancer development, such as Helicobacter pylori and human papilloma viruses [101].

Not only infections of pathogenic agents but also hosting species that live in symbiosis in the human body have shown growing evidence of their relationship with cancer, prompting scientific interest in recent decades. Indeed, there are studies estimating that in 2018, around 13% of global cancer incidence could be attributed to infection [102], as well as studies that relate microbial dysbiosis and cancer [103].

In recent years, microbiome dysbiosis has sparked interest in hematological diseases [104], and some studies evaluate the influence of microbiota on efficacy and treatment responses, including HSCT [105,106]. As certain bacterial species from the human microbiome secrete pathogenic products, some secreted molecules could provoke cell apoptosis, immune defense evasion, inflammatory processes or cancer evolution [107]. Therefore, microbiota composition could be responsible for patients’ inflammation state. As inflammation has been widely reviewed as a driver of MPN development [96], the dysbiosis of species related to this process could be particularly important in the disease context.

Specifically referring to MPNs, only a few studies have evaluated microbial dysbiosis. A first approach analyzed the microbial content in the PB and BM of 1870 newly diagnosed patients with myeloid malignancies (354 MPN patients) and compared it to 12 healthy controls [108]. Using deep DNA sequencing, the authors catalogued the bacterial, fungal and viral content in circulation, discovering dysbiosis in disease cases and different microbial fingerprints. The main differences were found in AML and MDS entities. Concretely in MPNs, the authors only suggested some kind of relationship between JAK2 mutation presence and the circulating microbiome, but further studies are required. Either way, this study served as a baseline for future research analyzing microbiomes in myeloid malignancy patients.

As the connection between BM and the gut is well established [109], other researchers have also evaluated stool samples in their studies. Barone and colleagues [110] isolated the microbial DNA cargo of circulating MK- and PLT-derived extracellular vesicles due to their role in the inflammatory network [111], together with fecal samples in PV patients’ samples and carried out 16S rDNA V3-V4 region sequencing. They found a higher diversity and different composition of microbial content, with a potential role in inflammation, in extracellular vesicles from PV than in healthy controls. Although they could not identify a differential microbiome profile of the gut microbiota of PV and controls, they associated it with the cytotoxic therapy administered during the time of the study [110].

In contrast, the Canadian group lead by Fleischman [112,113] compared the gut microbiota of triplicate fecal samples from 25 MPN patients (PV, ET and secondary MF) and 25 healthy controls (co-inhabitants whenever possible). The researchers found 1.7% of microbial composition variance explained by the disease. In this sense, they detected lower reads of species, such as Phascolarctobacterium, associated with reduced inflammation, as well as higher levels of the Prevotellaceae family linked to chronic inflammatory conditions in MPNs. These results support the important role of inflammation as one of the main promoters of the disease.

It is important to mention that most of the previous described studies [110,112,113] include patients under treatment that could modulate microbiome content, explaining some of the differences between studies. In addition, microbiome content is very individual and specific, and it changes according to diet and other external factors; therefore, establishing the correct controls is always challenging to obtain the best results.

Other researchers have focused on the evaluation of the implication of environmental exposures in normal population driving to MPN. In this line, recent studies utilizing data from “The Danish Health Examination Survey” (DANHES) have associated smoking with MPN development. Among 75,896 included patients, 70 were newly diagnosed with MPNs (41 women and 29 men). The analysis identified smoking as a significant risk factor for the development of MPNs compared to patients in the general population in Denmark who had never smoked [114]. Indeed, it is known that smoking is a risk factor, especially for PV development [115], because it affects inflammation and oxidative stress [116].

In a similar approach, two different population-based cohorts, the Copenhagen General Population Study and the Copenhagen City Heart Study, were used to evaluate HDL levels and the cancer risk of several entities [117]. The authors identified that low HDL cholesterol and/or apolipoprotein A1 were associated with an increased risk of MPNs (HR = 1.66) and other malignancies, such as multiple myeloma, non-Hodgkin lymphoma, breast cancer, lung cancer and nervous system cancer [117].

Likely, one of the main mechanisms driving to MPNs is inflammation imbalance. Alterations in the process occur in both malignant and non-malignant clones as well. Therefore, not only intrinsic factors from MPN malignant cells but also tumor microenvironments might provoke a chronic inflammatory state that may eventually lead to the development of MPNs [96,118].

Nonetheless, other mechanisms not so widely known in MPNs, which are also important, are oxidative stress and cellular metabolism derived from malignant cells and tumor microenvironments [119]. Oxidative disturbances and metabolic alterations display a considerable position in tumorigenic processes [120,121], causing alterations in the normal cell fate decision processes.

The oxidative status has been evaluated in PV, ET [122,123] and MF patients [124]. As with other myeloid malignancies, higher ROS levels and lower antioxidant capacity were observed in MPN patients compared to controls.

Metabolic dysregulation has also been well established in myeloid neoplasms, such as MDS [125] and AML [126], where the survival of leukemic stem cells (LSCs) has been explained by the alteration of different sources for energy production [127,128]. In MPNs, mutant JAK2 has been described to enhance the expression of glycolytic enzymes that sustain cell growth [129]. In vivo experiments have shown that JAK2-mutant hematopoietic cells displayed metabolic alterations essential for the pathogenesis of these neoplasms, leading to hypoglycemia, adipose tissue atrophy and early mortality [130] that could be reversed in the murine model using 3-(3-pyridinyl)-1-(4pyridinyl)-2-propen-1-one, which inhibits Pfkfb3, a key regulator of glycolysis. A similar approach in mice carrying both NRAS and EZH2 mutations showed that the hyperactivation of branched-chain amino acid (BCAA) metabolism due to the aberrant activation of BCAT1 was responsible for more aggressive MPNs with rapid progression to the acute leukemia phase, suggesting dietary BCAA restriction or the use of BCAT1 inhibitors in EZH2-mutated myeloid neoplasms [131]. These results reinforce the potential for targeting metabolism in mutant MPN cells [132].

On the other hand, the interaction between malignant and non-malignant clones in MPN patients showed different metabolic profiles in PV and ET patients compared to healthy controls, reflecting the energetic demands for fast proliferation in MPN patients [133]. Nevertheless, further studies on metabolism dysregulation in MPNs are needed to determine whether specific targeted drugs may be useful as a novel treatment for the disease [10].