Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

Comparison of whole blood vs purified blood granulocytes for the detection and quantitation of JAK2V617F

The V617F mutation of JAK2 (JAK2V617F) is specific of myeloproliferative disorders (MPDs) and its detection is now one of the first tests performed in the diagnostic work-up of polycythemia and thrombocytoses.1, 2, 3 The JAK2V617F allele burden varies depending on the type of MPD, with the majority of polycythemia vera (PV) expressing >50% JAK2V617F and essential thrombocythemia (ET) expressing <50% mutant4, 5 in purified granulocytes, the source of DNA typically used to study the JAK2V617F allelic ratio. However, granulocyte purification being time-consuming, many diagnostic laboratories opt to work on whole blood instead. There are potential problems with using DNA extracted from whole blood rather than purified granulocytes. B and T lymphocytes typically do not express JAK2V617F, which implies that the JAK2V617F burden is likely to be underestimated in whole blood assays. Moreover, some PV and half of JAK2-mutated ET express <15% JAK2V617F: such low levels of mutant, often missed by sequencing, are detected in purified granulocytes by sensitive allele-specific PCR (AS-PCR) techniques. In whole blood assays however, some of the patients with low levels of mutant might be found negative.

We used a allele-specific, quantitative PCR (AS-qPCR) that detects 0.1% JAK2V617F in granulocytes5, 6 to compare results obtained with genomic DNA extracted from whole blood or purified blood granulocytes. Forty-eight consecutive patients, 41 with MPDs (12 PV, 26 ET and three idiopathic myelofibrosis (IMF)) and seven with secondary erythrocytosis (SE), were examined. Twelve MPD patients (five PV, five ET and two IMF) had been receiving cytoreductive treatment (pipobroman or hydroxyurea); all other patients were examined at the time of diagnosis. Aliquots of whole blood and pellets of purified granulocytes were prepared from the same sample, as described5, 7 and stored at −20°C (whole blood) or −80°C (granulocyte pellets) until use. Whole blood DNA and granulocyte DNA were prepared using QiaAmp DNA mini-kits (Qiagen) following the manufacturers' instructions. JAK2 AS-qPCRs were performed at least in duplicate with 25 ng DNA per reaction, as described previously;5, 6 copy numbers were determined by comparison with serial dilutions of plasmids containing wild-type and V617F JAK2 DNA amplicons.

The JAK2 status (JAK2V617F-positive or -negative) of patients was identical in whole blood and in purified granulocytes: 36 MPD patients were found positive for JAK2V617F including the two patients with 1% mutant (Table 1); all SE and five ET were JAK2V617F-negative. A good correlation was found between the % JAK2V617F found in whole blood and the % JAK2V617F measured in purified granulocytes (n=36, r=0.98, P<0.001). The JAK2V617F allelic ratio was 15% lower on average in whole blood than in granulocytes; underestimation of the % JAK2V617F was similar in PV (−18%) and ET (−11%, P=0.320). MPD patients had on average 19% lymphocytes (JAK2V617F-negative cells), <10% monocytes and 70% neutrophils in whole blood whereas granulocyte preparations contained on average 86% neutrophils and <5% lymphocytes and monocytes, the rest being eosinophils (Table 1). Thus, the degree of underestimation of the JAK2V617F allelic ratio measured in whole blood likely depends on the % neutrophils. Indeed, when the % JAK2V617F observed were corrected for the % neutrophils in the DNA preparation, % JAK2V617F in whole blood were no longer underestimated: compared to purified granulocytes, the median variation of the % JAK2V617F in whole blood was 0% (corrected) vs −14% (observed).

Table 1 The JAK2V617F allelic ratio in whole blood and in granulocytes in relation to the % neutrophils.

We then analysed the effect of cytoreductive drugs on the assessment of the JAK2V617F allele burden. For the 24 untreated patients (seven PV, 16 ET and one IMF), the % JAK2V617F in whole blood correlated with the % JAK2V617F in purified granulocytes (n=24, r=0.99, P<0.001) (Figure 1a). The % JAK2V617F in whole blood also correlated with the white blood cell (WBC) counts (n=24, r=0.83, P<0.001), as reported previously5 and with the % neutrophils in blood (n=24, r=0.51, P=0.01). The 12 patients with cytoreductive treatment (five PV, five ET, two IMF) were characterized by lower WBC counts (6.4±2.7 × 109/l vs 11.7±6.4, P=0.010), a lower % neutrophils (62 vs 74% P<0.001) and a higher % lymphocytes (24 vs 17%, P=0.006) (Table 1). For treated patients, the % JAK2V617F in whole blood and the % of neutrophils were no longer correlated. This was not due to the small size of the series, as correlations persisted for groups of 12 patients selected randomly among patients at diagnosis. However, the % JAK2V617F in whole blood and the % JAK2V617F in granulocytes were correlated (n=12, r=0.95, P<0.001) (Figure 1b). The average underestimation of JAK2V617F burden in whole blood was – 8% for patients at diagnosis but – 29% for treated patients (P=0.002). When the % JAK2V617F observed was corrected by the % neutrophils in the DNA preparation, % JAK2V617F in whole blood and % JAK2V617F in granulocytes remained well correlated (n=24, r=0.99, P<0.001 for patients at diagnosis; n=12, r=0.97, P<0.001 for treated patients). However, in contrast to patients at diagnosis, the variation of the % JAK2V617F measured in whole blood of treated patients was not significantly reduced when adjusted for the % neutrophils in blood (Table 1). Reasons for the lack of correction are unclear. Treated patients had high % of both lymphocytes and monocytes (75% had >10% monocytes in blood, vs 25% of patients at diagnosis); unlike lymphocytes, monocytes carry JAK2V617F but they have been shown to express the mutation less frequently than granulocytes.8 Several of the treated patients also had an abnormally high % of eosinophils, the level of JAK2V617F expression of which is unknown (Table 1). Last, although there is so far no clear evidence of it, the possibility that mutated granulocytes may be more sensitive to the cytoreductive effect of hydroxyurea and pipobroman than non-mutated granulocytes cannot be excluded.

Figure 1

Correlations between the % JAK2V617F observed in whole blood and the % JAK2V617F observed in purified granulocytes. (a) Patients at diagnosis; (b) patients with cytoreductive treatment.

In summary, provided that the AS-qPCR used is sensitive (i.e., detects <1% JAK2V617F), there is no false negative when JAK2V617F is detected in whole blood. Hence, whole blood is suitable for the diagnosis of MPD including those with as little as 1% mutant; the JAK2V617F burden thus measured is lower than in purified granulocytes and may be adjusted using the % neutrophils. For the follow-up of the mutated clone in treated patients, however, it is advisable to assess JAK2V617F before and during treatment in the same population of purified cells as blood cell ratios are altered and the variation of the JAK2V617F burden measured in whole blood increases and no longer correlates with blood parameters. For the purpose of assessing the JAK2V617F burden during treatment, blood granulocytes remain the easiest cells to obtain and purify.


  1. 1

    James C, Ugo V, Le Couedic JP, Staerk J, Delhommeau F, Lacout C et al. A unique clonal JAK2 mutation leading to constitutive signalling causes polycythaemia vera. Nature 2005; 434: 1144–1148.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  2. 2

    James C, Delhommeau F, Marzac C, Teyssandier I, Couedic JP, Giraudier S et al. Detection of JAK2 V617F as a first intention diagnostic test for erythrocytosis. Leukemia 2006; 20: 350–353.

    CAS  Article  Google Scholar 

  3. 3

    Girodon F, Lippert E, Mossuz P, Dobo I, Boiret-Dupré N, Hermouet S, Praloran V . JAK2V617F detection and dosage of serum erythropoietin: first steps of the diagnostic work-up for patients consulting for elevated haematocrit. Haematologica, in press.

  4. 4

    Passamonti F, Rumi E, Pietra D, Della Porta MG, Boveri E, Pascutto C et al. Relation between JAK2 (V617F) mutation status, granulocyte activation, and constitutive mobilization of CD34+ cells into peripheral blood in myeloproliferative disorders. Blood 2006; 107: 3676–3682.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  5. 5

    Lippert E, Boissinot M, Kralovics R, Girodon F, Dobo I, Praloran V et al. The JAK2-V617F mutation is frequently present at diagnosis in patients with essential thrombocythemia and polycythemia vera. Blood 2006; 108: 1865–1867.

    CAS  Article  Google Scholar 

  6. 6

    Boissinot M, Lippert E, Girodon F, Dobo I, Fouassier M, Masliah C et al. Latent myeloproliferative disorder revealed by the JAK2-V617F mutation in patients with splanchnic vein thrombosis. Blood 2006; 108: 3223–3224.

    CAS  Article  Google Scholar 

  7. 7

    Dobo I, Donnard M, Girodon F, Mossuz P, Boiret N, Boukhari R et al. Standardization and comparison of endogenous erythroid colony assays performed with bone marrow or blood progenitors for the diagnosis of polycythemia vera. Hematol J 2004; 5: 161–167.

    CAS  Article  PubMed  Google Scholar 

  8. 8

    Kralovics R, Teo S-S, Li S, Theocharides A, Buser AS, Tichelli A, Skoda RC . The V617F mutation of JAK2 is a late genetic event in a subset of patients. Blood 2006; 108: 1377–1380.

    CAS  Article  PubMed  Google Scholar 

Download references

Author information



Corresponding author

Correspondence to S Hermouet.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Hermouet, S., Dobo, I., Lippert, E. et al. Comparison of whole blood vs purified blood granulocytes for the detection and quantitation of JAK2V617F. Leukemia 21, 1128–1130 (2007).

Download citation

Further reading


Quick links