Laboratory tests for Polycythemia vera 2



Laboratory tests for Polycythemia vera

JAK2 V617F and Exon 12 MutationsJAK2 G1849T Creating V617F

Mutations within the JAK2 gene cause a deregulation of the hematopoietic process, which is expressed in a wide spectrum of disorders involving the expansion of erythrocytes, granulocytes, or leukocytes. The primary lesion associated with these disorders was discovered in exon 14 and is a single nucleotide change JAK2 G1849T creating V617F. The detection of the JAK2 V617F mutation provides a qualitative diagnostic marker for the identification of Philadelphia-chromosome-negative subgroup of myeloproliferative disorders and their differentiation from congenital and acquired reactive hematopoietic disorders. In general, PMF patients have the highest, PV patients an intermediate, and ET patients the lowest JAK2 V617F allelic burden. In virtually all PV JAK2 V617F-positive patients at least some progenitors exist that became homozygous for JAK2 V617F mutation by uniparenteral disomy acquired by mitotic recombination. Allele-specific polymerase chain reaction (PCR) is widely used for single nucleotide polymorphism (SNP) genotyping, and the technique is based on amplification of DNA by an allele-specific primer matching the polymorphism at the 3′ position. This approach is directly applicable to analysis of JAK2 V617F because the mutation (G1849T) is analogous to a SNP. In theory, the allele-specific primer containing the mismatched nucleoside at the 3′ end should not be extended by Taq DNA polymerase. However, this DNA polymerase can extend mismatched allele-specific primers, generating false-positive results and decreasing sensitivity of mutation detection. To improve the specificity, sensitivity, and reliability, an allele-specific quantitative PCR method for quantitating the JAK2 V617F allele burden, two modifications into the basic technique were incorporated: (1) inclusion of a second mismatch at the –1 positions; (2) substitution of a modified locked nucleic acid (LNA) at the –2 position.34 A study comparing 11 different techniques was undertaken and carried out in 16 laboratories using various equipment.73 Although 5 of the 11 techniques were similarly reliable for the quantification of JAK2 V617F loads ≥1 percent of total JAK2, the allele-specific quantitative PCR technique could detect 0.2% JAK2 V617F.

The majority of laboratories analyze the JAK2 V617F allele burden from (clonal) granulocytes; the nonquantitative analyses use total leukocytes, whole blood, or marrow for screening. A proportion of JAK2 V617F-negative assays are positive using sensitive quantitative analyses.73 Plasma has been used for detection of the JAK2 V617F DNA and mRNA mutation and zygosity state.74,75 Detection of JAK2 V617F DNA and mRNA in plasma is, however, the result of the lysis of granulocytes during storage, compounded by decreased viability of the JAK2 V617F-bearing cells and is a storage artifact76; as such it should not be used in clinical practice to detect, quantify JAK2 V617F, or address zygosity.

JAK2 Exon 12 Mutations

Although the most common JAK2 mutation is a single SNP in exon 14, many MPD cases negative for the exon 14 mutation JAK2 V617F carry one of a number of mutations in the 3′ terminus of exon 12. These mutations occur within codons 537 to 544 and consist of single and multiple base substitutions and small deletions. Ten different mutations comprising base substitutions have been identified thus far with small base deletions and sequence duplication through this region.77–81 These mutations have been observed primarily in patients with isolated erythrocytosis. In addition to the varied types of mutations observed in this region, the proportion of mutation within a sample may be small and therefore difficult to detect in a high background of a normal sequence.

PCR amplification in the presence of a short blocking oligonucleotide homologous to exon 12 (codons 537–544) of the wild-type JAK2 gene. The oligonucleotide is designed to specifically suppress PCR amplification of the wild-type JAK2 exon 12 sequence. In contrast, JAK2 exon 12 mutations located between JAK2 codons 537 and 544 disrupt proper binding of the blocking oligonucleotide during PCR amplification, resulting in a product of approximately 225 base pairs. Each assay includes control DNA from mutation-positive and wild-type–negative samples; all samples are tested in paired reactions with and without blocking oligonucleotide. When a PCR product is formed solely in the presence of a blocking oligonucleotide, the sample is suspected of harboring a mutation and is sent for confirmatory sequencing. An alternative approach has been to use one of the modifications of the high-resolution melting analysis.77

A V617F-negative patient may have PV and require a search for other JAK2 mutations, or they may have another type of polycythemia (see Chap. 56).

V617-positive patients may have another MPD and require additional studies (see Chaps. 87, 91).

Marrow Findings

The marrow is characteristically hypercellular with an increase in erythroid and granulocytic precursor cells and megakaryocytes. Whereas marrow morphology is part of a World Health Organization (WHO) diagnostic criteria of PV,82 the morphologic features have not yet been validated and may be subject to inter- and intraobserver variations. Unpublished data suggest that the marrow in PV patients with exon 12 mutations is morphologically normal.83 Marrow morphology in the related myeloproliferative disorder ET was unreliable as a criterion for diagnosis.84 Absent or decreased iron stores are seen in the marrow of most PV patients. There are no characteristic cytogenetic findings, but occasional clonal chromosomal changes, none of which is very specific for PV, are observed in a minority of patients (see Chap. 11).

Blood Findings


The hemoglobin concentration, erythrocyte count, and hematocrit are usually increased and the mean cell volume is usually low-normal or low in untreated patients; in patients who have undergone phlebotomy or who have had gastrointestinal bleeding episodes, the erythrocyte count may be increased disproportionately to the increase in the hemoglobin and hematocrit, resulting in marked hypochromia and microcytosis. The plasma iron in such patients is decreased, the iron-binding capacity increased, and plasma ferritin levels are low. The red cell mass is usually increased in proportion to the hemoglobin concentration (see Fig. 86–2). However, because of the expanded plasma volume, the hemoglobin may be normal.85 In some series, the hemoglobin was falsely lower compared to the red cell mass without any apparent reason.86,87 Aniso- and poikilocytosis and teardrop cells herald the onset of the spent phase.

Red Cell Mass Determination

The polycythemia vera study group employed the direct determination of the red cell mass as the sine qua non of the diagnosis of PV in patients entered into their studies.48 Some believe that even in the routine clinical setting, this procedure should be performed on all patients to establish this diagnosis.48,88 Unfortunately, the determination of the red cell mass is expensive, requires the use of radioactive isotopes in patients, and, when performed by the inexperienced, is often inaccurate.89 It is not useful in distinguishing PV from secondary polycythemia, the differentiation that is usually needed, because it is increased in both disorders. The principal value of a red cell mass determination might then be to distinguish apparent or spurious polycythemia from PV and secondary polycythemia, since the elevated red blood cell mass was masked by simultaneous elevation of plasma volume.86,87 It may be useful in distinguishing some cases of “hidden PV” from cases of essential thrombocythemia.85 Ideally, the red cell volume and plasma volume should be measured separately. The availability of the JAK2 assay has made the blood cell mass determination only very rarely important.

The effect of the elevated red cell mass on blood viscosity is discussed under “Treatment” below.


An absolute neutrophilia occurs in about two-thirds of the patients.45 Occasional myelocytes and metamyelocytes are present in the blood, and considerable degrees of immaturity are present in patients with long-standing, advanced disease. Again, these abnormalities herald the onset of the spent phase (see Chap. 91). Basophilia occurs in approximately two-thirds of patients with uncontrolled disease.28,51,90 In PV, the proportion of activated neutrophils is increased91 and it is possible that neutrophils may be an important factor in PV-associated thrombosis.

The leukocyte alkaline phosphatase level is elevated in approximately 70 percent of patients with PV,45 but this assay has now become largely obsolete.45


The platelet count is increased in approximately 50 percent of patients at the time of diagnosis, and in approximately 10 percent if it is greater than 1000 x 109/L.45 In contrast to normal individuals in whom phlebotomy results in an increase in the platelet count, platelet levels may not be affected by phlebotomy in patients with PV.92 There are no consistent abnormalities of thrombopoietin levels.93 A significant proportion of PV patients first present with isolated thrombocytosis without an elevated hemoglobin and are sometimes misdiagnosed as having essential thrombocythemia.94

Qualitative abnormalities of the platelets have been described. In vitro spontaneous platelet aggregation is accelerated. On the other hand, patients with PV, ET, and other MPDs have a nearly pathognomonic defect in the primary wave of platelet aggregation induced by epinephrine.95 In contrast, there is increased platelet thromboxane A2 generation96 and increased excretion of thromboxane metabolites,97 even though the response to thromboxane A2 may be subnormal.98 Platelet factor-4 levels are elevated99 and platelet survival is normal100 or shortened.92,99 Fibrinogen binding after stimulation with a platelet activating factor is diminished101 and there is reduced expression of the thrombopoietin receptor.17 However, none of these so far described changes is specific for PV. Platelet counts over 1 to 1.5 millions are associated with progressive decrease of von Willebrand factor and increase risk of bleeding but not thrombosis.102

In a prospective study, the PIA2 polymorphism of the platelet glycoprotein (GP) IIIa was associated with an increased risk of arterial thrombosis in PV patients.103 However, polymorphisms of GPIb and GPIa, or the presence of the prothrombin G20210A mutation or factor V Leiden mutation, did not correlate with thrombohemorrhagic events.103


Serum lysozyme levels are slightly increased in some patients,104 and because of the increased leukocyte turnover and increased B12 binding protein, the levels of vitamin B12 are usually increased.105 Hyperuricemia, a consequence of hyperproliferative myelopoiesis, is frequently encountered.28

Differential Diagnosis

Also refer to Chap. 33, Table 33–2, and Chap. 56, Figure 56–6.

The diagnostic task has been greatly facilitated by the discovery of the JAK2 V617F mutation that is present in 95% or more of all PV patients.

When a patient presents with polycythemia, the initial step should be a repeat of the blood counts as the hemoglobin concentration may reflect a transient decrease of plasma volume (spurious polycythemia). If the hemoglobin concentration is persistently elevated, hypoxia should be considered as a possible cause. An arterial oxygen saturation level (SaO2) of <92% suggests cardiac or pulmonary etiologies. However, in PV, the partial pressure of oxygen of the arterial blood (PaO2) is often slightly decreased,106 complicating the differential diagnosis.

When hypoxia is ruled out, determining whether the increased hemoglobin is acquired or congenital and whether there are other family members involved is essential. The following tests should be employed, depending on availability and other specific circumstances to aid with the diagnosis of PV: (1) The JAK2 V617F mutation assay (see JAK2 V617F AND EXON 12 MUTATIONS) (2) Complete blood counts. In the 5 percent of PV patients without the JAK2 V617F mutation, the most important diagnostic features of PV are erythrocytosis, leukocytosis (especially basophilia or eosinophilia), thrombocytosis, and splenomegaly. Frequently, only two or three of these features are found at presentation and, if sufficiently pronounced, suffice to establish (at least preliminary) diagnosis. In some patients, only one of these features is found initially, most commonly erythrocytosis, but occasionally only thrombocytosis94 and, less often, leukocytosis or splenomegaly. Such patients represent more difficult diagnostic challenges. A subset of patients with erythrocytosis do not develop the other features of PV, even after they have been followed for many years.107,108 Such patients have been designated as manifesting pure erythrocytosis or idiopathic erythrocytosis.109 Some of these PV patients will have exon 12 JAK2 mutations.78,81 In JAK2 V617F negative patients: (3) Serum erythropoietin (EPO) level is usually low (see Fig. 56–6). In a small proportion of PV patients the EPO level may be normal. Such cases are often found in the presence of Budd-Chiari syndrome, iron deficiency, or postphlebotomy. (4) If the EPO level is normal or elevated, the P50 level (partial pressure of oxygen in the blood at which 50% of the hemoglobin is saturated with oxygen) should be measured (see Chaps. 48, 49 and 56). Polycythemic disorders resulting from high-affinity hemoglobin mutants (see Chap. 48) or low 2,3-bisphosphoglycerate (2,3-BPG) concentrations (see Chap. 46) are diagnosed with a decreased P50 from a Hemox-Analyzer, an instrument that records blood oxygen equilibrium curves. If a Hemox-Analyzer is not available, the P50 value can be calculated from freshly obtained venous blood gasses using a formula that can be calculated in Excel computer software.110 (5) Red cell and plasma volume studies (see Red Cell Mass Determination) to rule out spurious polycythemia (see Chaps. 33 and 56) caused by chronic contraction of plasma volume (Gaisbock syndrome) and unsuspected polycythemia in the face of normal hemoglobin. (6) If there is no evidence of congenital and familial history (see Chap. 56), search for exon 12 JAK2 mutations.

Distinguishing between PV and other polycythemic disorders may, at times, be challenging. Although the diagnosis of PV may be straightforward if patients have the classic criteria as defined by the most recent WHO criteria,82 (Table 86–1) patients often present with an incomplete phenotype. Some of the clinical and laboratory features that can be helpful for differential diagnosis are summarized in Table 33–2 and in Fig. 56–6. The current WHO diagnostic criteria, presented in Table 86–1,82 represent an improvement over the previous ones, although they do not necessarily discriminate between individual MPDs111 and they are yet to be validated by clinical studies. Children with PV are especially unlikely to fit the most recent WHO criteria.112

Table 86–1. World Health Organization Criteria for the Diagnosis of Polycythemia Vera, 2008
Major Criteria Minor Criteria
A1 Hgb >18.5 g/dL (men) >16.5 g/dL (women) or Hgb >17 g/dL (men), or >15 g/dL (women) if associated with a sustained increase of ≥2 g/dL from baseline that cannot be attributed to correction of iron deficiency Marrow trilineage myeloproliferation
A2 Presence of JAK2 V617F or similar mutation Subnormal serum EPO level
EEC growth
EEC, endogenous erythroid colony; EPO, erythropoietin; Hgb, hemoglobin.Both major and 1 minor or first major and 2 minor.

Erythroid Colony Cultures

In vitro assays of erythroid progenitor cells permit the study of their responsiveness to erythropoietin. In PV, erythroid BFU-E progenitors grow in serum containing cultures without added erythropoietin14 and the colonies are referred to as EECs. Detection of EECs in cultures of blood or marrow or blood may be the most specific test for PV.13,14,113,114 In one study, all patients with PV but none with secondary or other causes of polycythemia formed EECs in vitro.115 However, rare EECs may, at times, be observed in primary and familial polycythemia (PFCP) and in Chuvash polycythemia, but unlike the EEC of PV, these are abrogated by pretreatment with erythropoietin and erythropoietin receptor-blocking antibodies.116,117

In experienced hands, the EEC assay is a specific and sensitive means for detecting PV and may be useful in diagnosing patients with unusual presentations of PV, such as Budd-Chiari syndrome,53,55,118,119 isolated thrombocytosis, or in V617F-negative patients. In the era of JAK2V617F, this test, which has not been standardized and is expensive and laborious, is useful primarily in a research setting where it remains the gold standard for the diagnosis of PV.

Studies of red cell progenitors suggest that patients who have been diagnosed as having pure erythrocytosis can be divided into two groups of about equal size, those with erythropoietin-independent BFU-E and those without such precursors.107,115 It is possible that EEC negative pure erythrocytosis is a distinct entity, but those with EEC should be considered to have PV.

Erythropoietin Levels

Because PV is distinguished by the fact that erythroid cells proliferate even in the absence of normal levels of erythropoietin, one would expect that at high hematocrit levels the production of EPO would be reduced and the serum EPO levels would consequently be low. Several studies have indeed documented serum EPO levels below the normal reference range in patients with PV.120-122 In contrast to normal individuals, the serum EPO levels remain low even after phlebotomy.120 Patients with secondary polycythemia usually have normal to elevated erythropoietin levels, although considerable overlap exists in the range of erythropoietin levels between patients with PV and those with secondary polycythemia.121,123 Although an elevated EPO level generally excludes the diagnosis of polycythemia vera, a low EPO level is not pathognomonic of PV. Patients with PFCP also have as low EPO levels.124 Some PV patients with exon 12 JAK2 mutations were noted to have normal serum EPO levels.35

Clonality in Female Subjects Using Assays Employing X-Chromosome–Based Polymorphism Assays

PV results from an acquired mutation in a pluripotential hematopoietic cell. Clonality studies based on the phenomenon of X-chromosome inactivation125 show that red cells, granulocytes, platelets, monocytes, and B lymphocytes are all part of the neoplastic clone.12,13,126 The majority of T lymphocytes and natural killer cells are polyclonal, but a small proportion of these cells are also derived from the PV clone.11 This is presumed to be a result of the detection of the presence of long-lived normal T cells that preceded the development of the clone. Unfortunately, the applicability of X-chromosome inactivation for the differential diagnosis of PV is hampered by the many methodologic and conceptual differences that have drawn conflicting conclusions. Some of discrepancies are a result of two different approaches comparable127 that are used to distinguish the active from the inactive X-chromosome; one using X-chromosome differential methylation,128 typically using the polymorphic CAG repeat in the human androgen-receptor gene,129 versus the more biologically sound but more technically demanding transcriptional analysis of the active X-chromosome130,131 (see Chap. 9). Furthermore, the wide range of skewing of the X-chromosome allelic usage that is normally present132 is often misinterpreted as monoclonality, and the potentially clonal myeloid cells are not compared to the polyclonal control cells of the same origin.13 In approximately 100 female PV patients, the reticulocytes, platelets, and granulocytes were always clonal with the exception of a few patients who converted to polyclonal hematopoiesis after therapy with interferon-.13

Other Proposed Tests of Proposed Diagnostic Utility

Assay for MPL

Thrombopoietin, the primary regulator of platelet production, is produced in the liver and its levels are regulated, in large part, by binding to its receptor, c-MPL. Levels of c-MPL on platelets and megakaryocytes are decreased in patients with PV.17 The major limitations of an assay of c-MPL levels in the diagnosis of PV are its difficulty and nonspecificity.11

Assay for bcl-x

Increased expression of bcl-x, an antiapoptotic gene, occurs in PV erythroid progenitors18 but this is not specific for PV.

Assay for Polycythemia Rubra Vera-1

Increased mRNA levels of a receptor termed polycythemia rubra vera-1 (PRV-1) have been reported in PV granulocytes but not their progenitors.20 The exact function of PRV-1 in normal hematopoiesis is unclear and likely plays no significant role in PV pathophysiology because there are no differences in the amount of this protein between normal and PV progenitors. Approximately 80 to 100 percent of PV patients have increased granulocyte PRV-1 mRNA.