Pathogenesis

PV is a clonal stem cell disease with trilineage myeloid involvement.¹³ In addition, some studies suggest clonal heterogeneity, including clonal involvement of B lymphocytes,¹⁴ as well as polyclonal granulopoiesis in certain cases.¹⁵ In vitro, erythroid colony formation in patients with PV does not require the addition of exogenous erythropoietin (Epo).¹⁶ This phenomenon is called endogenous erythroid colony growth and does not occur in either normal controls or in patients with nonclonal polycythemia. In addition, erythroid progenitor cells in PV display growth factor hypersensitivity to Epo, insulin-like growth factor (IGF)–1, and other cytokines.^17,18 The consistently observed IGF1 hypersensitivity of erythroid cells in PV has been attributed to alterations in IGF1 binding proteins.¹⁹ However, several studies show that growth factor–independent or growth factor–hypersensitive colony formation is specific to neither PV nor erythroid progenitor cells.¹⁸

The Epo receptor gene as well as its protein are intact in PV.²⁰ On the other hand, several postreceptor molecular abnormalities have been reported and include increased baseline phosphorylation of the IGF1 receptor,²¹ decreased activity of SH-PTP1 (a tyrosine phosphatase),²² increased activity of membrane-associated SH-PTP,²³ constitutive activation of STAT3,²⁴ upregulation of negative control elements of the cell cycle (p16/p14),²⁵ and abundance, in erythroid precursors, of antiapoptotic proteins (Bcl-x_L).²⁶ However, these observations have not been always reproducible, and none of them have been shown to be specific to PV, as opposed to other MPNs.

In 2005, a novel Janus kinase 2 (JAK2) mutation (JAK2V617F) was described in association with PV, ET, and PMF.²⁷ JAK2V617F occurs in approximately 95% of PV patients but also in approximately 50% of those with ET or PMF.^28–31 In addition, the mutation is found in a small proportion of patients with other myeloid neoplasms including nonclassic MPNs and MDS.^32,33 Other JAK2 mutations (i.e., JAK2 exon 12 mutations) were subsequently described in the majority of patients with JAK2V617F⁻ PV or “idiopathic erythrocytosis,”³⁴ thus raising the possibility that a JAK2 mutation is essential for the PV phenotype. On the other hand, neither JAK2V617F nor JAK2 exon 12 mutations have so far been reported in lymphoid disorders,^35–38 solid tumor,^41–41 or secondary myeloproliferation including congenital or acquired polycythemia.^44–44

JAK2V617F is a G-to-T somatic mutation, at nucleotide 1849, in exon 14, resulting in the substitution of valine by phenylalanine at codon 617.28–31 JAK2 exon 12 mutations (F537-K539delinsL, H538QK539L, K539L, N542–E543del) include both in-frame deletions and tandem point mutations.³⁴ Both exon 14 and exon 12 JAK2 mutations induce cytokine-independent/hypersensitive proliferation in Epo receptor–expressing cell lines and a PV-like phenotype in mice.³⁴ Whereas homozygosity for JAK2V617F, which results from mitotic recombination, can be demonstrated in the majority of patients with PV,⁴⁵ exon 12 JAK2 mutations are often heterozygous.³⁴ The precise pathogenetic role of JAK2 mutations in PV and related MPNs continues to be investigated.

Diagnosis

In clinical practice, the term polycythemia is used to indicate the possible occurrence of an increased erythrocyte volume or red blood cell mass (RCM). Such a perception might be either real (true polycythemia) or spurious (apparent polycythemia; Fig. 100-1).¹⁸ True polycythemia may represent either PV or a nonclonal increase in RCM that is often, but not always, mediated by Epo (secondary polycythemia; Fig. 100-2). Apparent polycythemia may result from either a reduction in plasma volume (relative polycythemia) or an inaccurate perception of an elevated RCM that results from not appreciating high normal values of hemoglobin/hematocrit.⁴⁶ Inapparent polycythemia is the converse of apparent polycythemia and indicates a true increase in RCM that is masked by a normal hemoglobin/hematocrit value secondary to a concomitant increase in plasma volume (see Fig. 100-1).⁴⁷

In 1975, the PV study group published a set of “diagnostic criteria” that were primarily used to ensure the exclusion, to treatment protocols, of patients with secondary or apparent polycythemia.⁴⁸ These criteria required the demonstration of increased RCM by blood volume measurement using labeled erythrocytes as well as the demonstration of normal hemoglobin oxygen saturation. In 2001, the WHO published a refined diagnostic criteria for PV and related MPNs that recognized the value of both bone marrow histology and “MPN-specific” biological parameters.⁴⁹ The recent discovery of the almost invariable association between PV and a JAK2 mutation has led to revised WHO diagnostic criteria for PV (Table 100-2)⁵⁰ and a refined diagnostic algorithm (see Fig. 100-2),⁵¹ both of which incorporate JAK2 mutation screening.

Because more than 95% of patients with PV carry the JAK2V617F mutation, which is absent in both secondary and apparent polycythemia, it is most effective to initiate the workup of a patient with suspected PV with peripheral blood mutation screening for JAK2V617F (see Fig. 100-2). Furthermore, to minimize the consequences of false-positive or false-negative molecular test results, as well as capture the few cases of PV that are JAK2V617F⁻, concomitant measurement of serum Epo level is recommended.⁵² If the results of both tests are suggestive of PV (i.e., mutation-positive and low serum Epo), then the diagnosis is likely and bone marrow examination is encouraged but not essential for making the diagnosis. If the JAK2V617F and serum Epo test results are both not consistent with the diagnosis of PV (i.e., mutation-negative and either normal or increased Epo), then further investigation for PV is not advised unless dictated otherwise by the clinical scenario.

If there is discrepancy between the molecular test result and serum Epo level, one should first repeat both tests and then proceed with bone marrow examination if the results are unchanged (see Fig. 100-2). In a patient with acquired erythrocytosis and a subnormal serum Epo level, the possibility of exon 12 JAK2 mutations should be entertained. In this regard it should be noted that some cases with JAK2 exon 12 mutation–associated PV might not display overt PV-characteristic features, even at the level of bone marrow histology.³⁴

When congenital polycythemia is suspected, initial laboratory testing should include measurement of the oxygen tension at which hemoglobin is 50% saturated (i.e., p50).⁵¹ Left-shifted oxygen dissociation curve, suggested by decreased p50, suggests the presence of either high oxygen-affinity hemoglobinopathy (autosomal dominant)⁵³ or 2,3-bisphosphoglycerate (2,3-BPG) deficiency, usually a consequence of BPG mutase mutation (autosomal recessive).⁵⁴ If the p50 is normal, then the possibility of germline mutations of molecules that either enhance Epo effect (e.g., EPOR mutations)⁵⁵ or govern intracellular oxygen sensing (e.g., mutations involving the von Hippel-Lindau [VHL] tumor suppressor or hypoxia-inducible factor 1a [HIF-1a] prolyl hydroxylase genes)^56,57 should be considered (see Fig. 100-2). VHL mutations are usually associated with increased serum Epo level and constitute the most frequent mutations in congenital polycythemia (e.g., Chuvash polycythemia).⁵⁸ In contrast, serum Epo level is often subnormal in patients with EPOR mutations.⁵⁹

Bone marrow histology, to the experienced hematopathologist, is often revealing of characteristic changes of a MPN that include hypercellularity, increased number of megakaryocytes including cluster formation, the presence of giant megakaryocytes and pleomorphism in megakaryocyte morphology, mild reticulin fibrosis, and decreased bone marrow iron stores.⁶⁰ In contrast, cytogenetic studies in PV disclose abnormalities (trisomies of chromosomes 9 and 8 and deletions of the long arms of chromosomes 13 and 20) in only 13% to 18% of patients at diagnosis, and hence have limited diagnostic value.⁶¹

Treatment

The natural history of PV is characterized by a lifelong propensity for thrombohemorrhagic complications, late-onset disease transformation into post-PV myelofibrosis and/or AML, and a shortened life expectancy.⁶¹ Specific treatment has been shown to positively influence the risk of both macrovascular and microvascular complications but not that of clonal evolution to post-PV myelofibrosis or AML.

Phlebotomy is the cornerstone of therapy in PV and is the only treatment modality that has improved survival in affected patients (Box 100-1). It is reported that median survival may be as low as 2 years in the absence of such treatment.⁶³ Based on limited retrospective studies in PV that showed a progressive increase in the incidence of vascular occlusive episodes above a hematocrit level of 44%,⁶⁴ as well as other studies that showed suboptimal cerebral blood flow in ranges of hematocrit values between 46% and 52%,⁶⁵ the therapeutic target hematocrit level has, for a long time, been set at or below 45%. This particular practice is now further supported by the results of a recent randomized study where 365 adults with PV were treated with a target hematocrit, less than 45% or 45% to 50%.⁶⁶ After a median follow-up of 31 months, the primary end point of thrombotic events or deaths from cardiovascular causes was recorded in 5 of 182 patients in the low-hematocrit group (2.7%) and 18 of 183 patients in the high-hematocrit group (9.8%) (P = 0.007). Furthermore, because of the physiological difference in hematocrit values between the two genders as well as among different races, it is reasonable, though not evidence-based, to target an even lower hematocrit level (i.e., 42%) in women and African-Americans.⁶⁷

Pathogenesis

Trilineage clonal myeloproliferation has been demonstrated in the majority of patients with ET using X chromosome–linked DNA or gene product analysis.104,105 However, X-linked clonal assays have revealed both polyclonal hematopoiesis in a substantial minority of patients with ET¹⁰⁶ and “monoclonal” hematopoiesis in normal elderly controls.¹⁰⁷ Furthermore, in some cases, the clonal process in ET was shown to include lymphocytes¹⁴ or be restricted to megakaryocytes.¹⁰⁵ The clonal basis of ET was recently underlined by JAK2V617F mutation analysis that revealed the presence of the mutation even in those patients who feature “polyclonal” hematopoiesis by X-linked clonality studies.^106,108 However, the primary clonogenic event in ET remains undefined despite the recent descriptions of two gain-of-function mutations that occur in 50% (JAK2V617F)¹⁰⁹ and 1% (MPLW515L/K)¹¹⁰ of patients with ET, respectively.

Other biological features in ET include in vitro growth factor independence/hypersensitivity of both erythroid and megakaryocyte progenitor cells,111,112 low serum Epo level,¹¹³ altered megakaryocyte/platelet Mpl expression,^114,115 increased neutrophil PRV-1 expression,^116,117 and decreased platelet serotonin content.¹¹⁸ Laboratory studies in ET have demonstrated myeloid growth factor hypersensitivity to interleukin-3¹¹⁹ as well as thrombopoietin (Tpo).¹¹² Growth factor independence of myeloid progenitor cells in ET and related myeloproliferative disorders has now been attributed to mutations involving molecules of the JAK-STAT pathway including the aforementioned JAK2V617F and MPLW515L/K. In patients with ET,¹²⁰ PV,¹²¹ and myelofibrosis,¹²² serum Tpo levels are usually normal or elevated despite an increased megakaryocyte mass. This has been attributed to the markedly decreased megakaryocyte/platelet expression of myeloproliferative leukemia (MPL), although the extent of this particular phenotypic abnormality is significantly less in ET as compared with PV or PMF.^{114,115,123,124}

In addition to clonal myeloproliferation, ET is also characterized by microvascular symptoms (e.g., headaches and erythromelalgia) as well as an increased risk of both thrombosis and bleeding.¹²⁵ Pathogenesis in the former might involve abnormal thromboxane A₂ generation and small vessel–based platelet-endothelial interactions that are inhibited by aspirin therapy.^126,127 On the other hand, there is increasing information that implicates granulocytes but not platelets as the thrombogenic culprit in ET.^88,128–130 Consistent with these observations, a recent study has identified leukocytosis as an independent risk factor for thrombosis in ET.¹³¹ Bleeding diathesis in ET is currently believed to involve an acquired von Willebrand syndrome that becomes apparent in the presence of extreme thrombocytosis.⁹⁰ The mechanism of acquired von Willebrand syndrome in ET is currently believed to involve a platelet count-dependent increased proteolysis of high-molecular-weight von Willebrand protein.⁹⁰ Other qualitative platelet defects in ET are believed to play a minor role in disease-associated hemorrhage and include defects in epinephrine-, collagen-, and adenosine diphosphate–induced platelet aggregation, decreased adenosine triphosphate secretion, and acquired storage pool deficiency that results from abnormal in vivo platelet activation.¹²⁵

Diagnosis

Although thrombocytosis is the hallmark of ET, more than 85% of cases with thrombocytosis seen in routine clinical practice are reactive (secondary thrombocytosis) and associated with other comorbid conditions (Table 100-6).¹³² The degree of thrombocytosis is a poor discriminator of ET from secondary thrombocytosis. However, the clinical scenario is often helpful in distinguishing ET from secondary thrombocytosis. In routine clinical practice, it is important to exclude the contribution of either iron-deficiency anemia or a hyposplenic state as possible causes of an otherwise unexplained thrombocytosis (Box 100-2). This is accomplished by the measurement of serum ferritin concentration and examination of a peripheral blood smear looking for Howell-Jolly bodies, respectively. The possibility of secondary thrombocytosis associated with an occult inflammatory or malignant process is addressed by the measurement of C-reactive protein or other acute-phase reactants.¹³³ In other words, an uncomplicated ET should be accompanied by a normal serum ferritin, mostly unremarkable peripheral smear, and a normal serum C-reactive protein. In general, plasma Tpo levels are not helpful in distinguishing secondary thrombocytosis from ET.¹²⁰

Unlike the case with PV, the utility of mutation screening for JAK2V617F for the diagnosis of ET is limited by suboptimal negative predictive value and lack of diagnostic specificity within the context of myeloid neoplasms. It should be noted that only half of patients with either ET carry JAK2V617F and that the presence of the mutation cannot differentiate ET from another MPN. Therefore although peripheral blood screening for JAK2V617F helps in streamlining further investigation, a bone marrow biopsy is often required to help with the differential diagnosis of primary thrombocytosis. In general, the presence of JAK2V617F argues against the possibility of a nonclonal process such as infection or inflammation or a nonmyeloid malignancy such as lymphoma or metastatic cancer.

The typical bone marrow finding in ET consists of a mild to moderate increase in cellularity and the presence of megakaryocyte clusters that are often absent in secondary thrombocytosis (Fig. 100-3). It should be remembered that CML and the cellular phase of PMF could both mimic ET in presentation.¹³⁴ With respect to CML, a peripheral blood or bone marrow fluorescence in situ hybridization study may be needed, in addition to cytogenetic analysis, so as to exclude the possibility of karyotypically occult CML.¹³⁵ Similarly, bone marrow histology should be carefully scrutinized for the presence of intense marrow cellularity with florid atypical megakaryocytic hyperplasia suggesting cellular phase PMF. Cytogenetic abnormalities are rare in ET (<5%) and diagnostically not helpful.¹³⁶ Taking all the preceding information into consideration, bone marrow examination remains central to the diagnosis of ET, and its importance is underlined in the newly revised WHO criteria for the diagnosis of ET (Table 100-7).

Treatment

The majority of patients with ET either are asymptomatic or suffer from non–life-threatening microvascular disturbances (headache, visual symptoms, lightheadedness, atypical chest pain, acral dysesthesia, erythromelalgia) that are effectively treated with low-dose aspirin.¹³⁷

Approximately 20% of patients experience, usually nonfatal, thrombohemorrhagic complications, and less than 5% progress into AML or post-ET myelofibrosis during the first decade of the disease.131,138–141 Therefore it is reasonable to expect long survival in the majority of patients with ET.

In an international collaborative study of 1104 patients with clinically suspected ET,¹⁴² diagnosis was confirmed as true WHO-defined ET in 891 patients (81%) and was revised to early/prefibrotic PMF in 180 (16%). In early/prefibrotic PMF compared with ET, the 10-year survival rates (76% and 89%, respectively) and 15-year survival rates (59% and 80%, respectively), leukemic transformation rates at 10 years (5.8% and 0.7%, respectively) and 15 years (11.7% and 2.1%, respectively), and rates of progression to overt myelofibrosis at 10 years (12.3% and 0.8%, respectively) and 15 years (16.9% and 9.3%) were significantly worse. Multivariable analysis confirmed these findings and also identified age older than 60 years leukocyte count greater than 11 × 10⁹/L, anemia, and thrombosis history as additional risk factors for survival. Furthermore, survival in ET was similar to the sex- and age-standardized European population. In analysis of the 891 patients with ET, after a median follow-up of 6.2 years, 109 (12%) patients experienced arterial (n = 79) or venous (n = 37) thrombosis. In multivariable analysis, predictors of arterial thrombosis included age more than 60 years, thrombosis history, cardiovascular risk factors including tobacco use, hypertension, or diabetes mellitus, leukocytosis (>11 × 10⁹/L), and presence of JAK2V617F.¹⁴³ In contrast, only male gender predicted venous thrombosis. Interestingly, a platelet count greater than 1000 × 10⁹/L was associated with a lower risk of arterial thrombosis.

The relatively low incidence figures of thrombosis and hemorrhage as well as the occurrence of both short-term and long-term drug side effects are the basis for carefully selecting the patients with ET who require specific treatment. Risk stratification in ET is similar to that of PV (see Table 100-4). Table 100-8 outlines risk-adjusted treatment guideline in ET that is further elaborated in the following sections.

Pathogenesis

In addition to clonal myeloproliferation, the bone marrow in PMF might display excess collagen fibrosis, new bone formation (osteosclerosis), and angiogenesis. These changes have been associated with alterations in both cellular and extracellular levels of various fibrogenic and angiogenic cytokines including transforming growth factor-β (TGF-β), basic fibroblast growth factor, and platelet-derived growth factor.¹⁵⁴ The demonstration of polyclonal fibroblast proliferation in PMF is the basis for the current assumption that the bone marrow stromal aberration in PMF is reactive¹⁵⁵ and mediated by the aforementioned cytokines derived from the resident clonal megakaryocytes and monocytes (Fig. 100-4).¹⁵⁶

Current evidence strongly supports the stem cell origin of the clonal myeloproliferation in PMF.157,158 However, the primary clonogenic mutation(s) has not been identified, although much attention has been given to recently described novel gain-of-function mutations involving the JAK2 tyrosine kinase (JAK2V617F) and Tpo receptor (MPLW515L/K).^31,159 JAK2V617F represents a G-to-T somatic mutation of JAK2, at nucleotide 1849, in exon 14, resulting in the substitution of valine to phenylalanine at codon 617.³¹ MPLW515L mutation represents a G-to-T transition at nucleotide 1544 resulting in a tryptophan-to-leucine substitution at codon 515 of the transmembrane region of the MPL receptor.¹⁵⁹

JAK2V617F occurs not only in PMF (~50% mutational frequency) but also in ET with a similar mutational frequency and in PV where it is present in more than 90% of affected patients.28–31 In contrast, MPLW515L/K seems to be specific to PMF or ET, although mutational frequency is substantially lower (~5% and 1%, respectively).¹¹⁰ As mentioned before, JAK2V617F induces constitutive JAK-STAT activation, cytokine hypersensitivity, and/or independence in cell lines, and a PV-like disease in mice.^31,160,161 JAK-STAT is similarly hyperactivated by MPLW515L/K, but this mutation induces a PMF-like disease in mice.¹⁵⁹ In any case, about half of the patients with PMF do not display either mutation, and the precise pathogenetic role of these mutations, when they are present, remains to be clarified. Furthermore, since the description of JAK2V617F in 2005, many other mutations involving LNK, CBL, TET2, ASXL1, IDH, IKZF1, EZH2, DNMT3A, TP53, SF3B1, or SRSF2 have been described in PMF and other MPNs. However, none of them appears to garner the disease specificity or pathogenetic relevance otherwise displayed by BCR-ABL1. Nevertheless, the phenotype of clonal erythrocytosis might require the presence of a mutation in JAK2 (JAK2V617F or JAK2 exon 12 mutation)¹⁶² or its negative regulators such as LNK.¹⁶³ Similarly, there appears to be a close association between SF3B1 mutations and presence of bone marrow ring sideroblasts in MPN.

An animal model of PMF has been established in mice that are either chronically overexposed to Tpo¹⁶⁴ or carriers of a mutant GATA-1 gene that results in reduced expression of a transcription factor that plays a role in erythroid and megakaryocyte differentiation.¹⁶⁵ These mice display characteristic features of PMF, including megakaryocytic hyperplasia, bone marrow fibrosis, osteosclerosis, and extramedullary hematopoiesis. The increased megakaryocyte accumulation in these experimental animals has been attributed to either a direct effect of Tpo or the lack of a negative feedback from mature elements as a result of impaired megakaryocyte differentiation, respectively. In both instances it is possible that the abnormal accumulation of megakaryocytes and their sequestered cytokines might be central to the pathogenesis of the associated stromal reaction. In this regard, the development of Tpo-induced bone marrow fibrosis in mice has been temporally associated with elevated TGF-β levels,¹⁶⁶ whereas it was abrogated in TGF-β knockout experiments.¹⁶⁷ Similar experiments may decipher the individual roles of other cytokines, including basic fibroblast growth factor and platelet-derived growth factor, that have been implicated in PMF.¹⁶⁸ It is currently not clear whether the aforementioned animal models of PMF involve pathogenetic mechanisms that are operating in the human form of the disease.

Diagnosis

The typical presentation of PMF includes anemia (from ineffective erythropoiesis), marked splenomegaly (from extramedullary hematopoiesis), and a myelophthisic peripheral blood smear. Myelophthisis (the presence of nucleated red blood cells, granulocyte precursors, and teardrop-shaped erythrocytes) suggests a bone marrow infiltrative process, and the differential diagnosis includes bone marrow fibrosis, metastatic cancer, granulomatous infection, and lymphoma (Box 100-3). In PMF, peripheral blood myelophthisis is associated with bone marrow megakaryocytic hyperplasia, collagen fibrosis, osteosclerosis, and intramedullary sinusoidal hematopoiesis (Fig. 100-5). It should be noted, however, that bone marrow fibrosis may accompany not only PMF but also other hematologic and nonhematologic disorders, including systemic mast cell disease, eosinophilic disorders, and hairy cell leukemia (Table 100-10). Therefore before embarking on treatment recommendations, one has to consult with an experienced hematopathologist to confirm a specific diagnosis. Extra caution should be taken not to misdiagnose CML or hairy cell leukemia as PMF. These issues are addressed and JAK2V617F mutation screening incorporated in the newly revised WHO diagnostic criteria for PMF (see Table 100-9).⁵⁰

PMF should be distinguished from other closely related myeloid neoplasms, including CML, PV, ET, MDS, CMML, and “acute myelofibrosis.” The presence of dwarf megakaryocytes raises the possibility of CML and should be pursued with BCR-ABL1 cytogenetic or molecular testing. Prefibrotic PMF can mimic ET in its presentation and careful morphologic examination is necessary for distinguishing the two; megakaryocytes in ET are large and mature-appearing whereas those in prefibrotic PMF display abnormal maturation with hyperchromatic and irregularly folded nuclei¹⁶⁹; the distinction between ET and prefibrotic PMF is prognostically relevant.¹⁷⁰ MDS should be suspected in the presence of dyserythropoiesis or dysgranulopoiesis.¹⁷¹ CMML is a possibility in the presence of peripheral blood monocyte count of greater than 1 × 10⁹/L. Patients with acute myelofibrosis (either acute panmyelosis with myelofibrosis or acute megakaryoblastic leukemia) usually present with severe constitutional symptoms, pancytopenia, mild or no splenomegaly, and increased circulating blasts.¹⁷²

Among the BCR-ABL1⁻ classic MPNs, PMF has the worst prognosis, with an approximate median survival of 5 years.¹⁷³ However, several studies have identified both clinical and laboratory parameters that are used to identify good-risk as well as high-risk patient categories.¹⁵¹ The most recently revised prognostic scoring system in PMF is called “DIPSS-Plus”¹⁷⁴ (Dynamic International Prognostic Scoring System–Plus) and can be applied at any point during the disease course and uses eight independent predictors of inferior survival: age older than 65 years, hemoglobin <10 g/dL, leukocytes >25 × 10⁹/L, circulating blasts ≥1%, constitutional symptoms, red cell transfusion dependency, platelet count <100 × 10⁹/L and unfavorable karyotype (i.e., complex karyotype or sole or two abnormalities that include +8, −7/7q−, i(17q), inv(3), −5/5q−, 12p- or 11q23 rearrangement). The presence of 0, 1, “2 or 3,” and ≥4 adverse factors defines low, intermediate-1, intermediate-2, and high-risk disease with median survivals of approximately 15.4, 6.5, 2.9, and 1.3 years, respectively. Furthermore, a greater than 80% 2-year mortality rate was predicted by monosomal karyotype (two or more autosomal monosomies or a single autosomal monosomy associated with at least one structural abnormality), inv(3)/i(17q) abnormalities, or any two of circulating blasts greater than 9%, leukocytes 40 × 10⁹/L or greater, or other unfavorable karyotype.¹⁷⁵ Patients with the latter characteristics are operationally assigned a “very high risk” category and might be better served by immediate consideration for allogeneic stem cell transplantation. More recent data suggest inferior survival in PMF associated with nullizygosity for JAK2 46/1 haplotype,¹⁷⁶ low JAK2V617F allele burden,^177,178 or presence of IDH,¹⁷⁹ EZH2,¹⁸⁰ SRSF2,¹⁸¹ or ASXL1¹⁸² mutations. Survival in PMF was also affected by increased serum IL-8 and IL-2R levels as well as serum-free light chain levels, both independent of DIPSS-Plus.^66,183,184

Risk factors for leukemia-free survival include 3% or greater circulating blasts, platelet count less than 100 × 10⁹/L, unfavorable karyotype, and presence of certain mutations, including IDH and SRSF2.179,181,185,186