The diagnosis of polycythemia vera (PV) is currently based on clinical and biological criteria defined either by the World Health Organization (WHO) or the Polycythemia Vera Study Group (PVSG). 1, 2 Both of these classifications use clinical and biological markers organized in major and minor criteria, allowing to diagnose a PV when a define combination of major and minor criteria is present. Recently, a somatic point mutation of the tyrosine kinase JAK2 (JAK2 V617F) has been described in about 80% of PV patients as well as 30% of essential thrombocythemia (ET) and 50% of idiopathic myelofibrosis (IMF) 3–6 and several in vitro and in vivo experiments demonstrated that this mutation of JAK2 was the molecular event at the origin of PV. 3 It was therefore tempting to use the detection of JAK2V617F as a diagnostic test for erythrocytosis. Such an attitude supposes the development of other techniques than sequencing, as this latter is time consuming and not always feasible in hematology laboratories. We thus compared sequencing with two techniques of real-time PCR-based mutation detection (one using the LightCycler® instrument, the other the Taqman® ABI Prism 7500), for the efficiency to detect the JAK2 V617F mutation in 119 samples from patients with suspicion of myeloproliferative disorder (MPD). The three techniques were equivalent in all samples but one, where sequencing failed to detect the mutation revealed by both LightCycler® and Taqman® technologies. To evaluate the sensitivity of these techniques, we tested serial dilutions of the homozygously mutated HEL cell line DNA in the nonmutated TF-1 cell line DNA, and serial dilutions of the genomic DNA from a patient homozygous for the JAK2 V617F mutation in normal DNA. Sequencing failed to detect the mutated allele under 5% of HEL cell line DNA diluted in TF-1 cell line DNA, and under 10% of the homozygously mutated patient’s DNA diluted in normal DNA. The sensitivity of LightCycler® and Taqman® techniques were equivalent, slightly higher than sequencing, reaching 0.5–1% of HEL cell line DNA diluted in TF-1 cell line DNA (Figure 1), and 2–4% of a homozygously mutated patient’s DNA diluted in normal DNA.

We then investigated in 88 patients with a newly diagnosed erythrocytosis if the detection of the mutation correlated with usual criteria of PV, as WHO and PVSG classifications, endogenous erythroid colonies (EEC) formation, and a low serum erythropoietin (Epo) level, thus facilitating the diagnosis of PV (Table 1). We retained 51% for the upper end of the normal range for hematocrit value. 7 Clinical data were incomplete for seven patients, explaining why the diagnosis of PV could not be affirmed either with the WHO or with the PVSG criteria. Owing to the differences between the A1 criteria of both classifications, six patients who did not have any red cell mass measurement could be classified in the WHO and not in the PVSG classification. One patient had both an hypoxia and EEC formation, therefore making the diagnosis difficult. A cytogenetic analysis was performed in 35 patients; among 32 PV patients (according to the WHO criteria), seven had cytogenetic abnormalities: five with trisomy 9, one with 7q-, and one with additional material on chromosome 18. JAK2 V617F was present in 43/45 (96%) patients diagnosed as PV according to the PVSG criteria and in 57/61 (93%) patients diagnosed with the WHO criteria (Table 1). Nevertheless, 8/29 patients classified as non-PV according to the PVSG classification presented the mutation, whereas none of the 19 non-PV patients of the WHO; these eight patients were considered