JAK2 exon 12 mutations in polycythemia vera or idiopathic erythrocytosis

Abstract

JAK2 exon 12 mutations were detected in 4 out of 20 polycythemia vera and idiopathic erythrocytosis V617F-negative patients and were only present in the myeloid lineage. Initial hematologic data of these patients differ from those of V617F-positive patients, but there is no difference in thrombotic development and myelofibrotic transformation.

Recently, several gain-of-function mutations affecting exon 12 of the JAK2 gene have been described in polycythemia vera (PV) V617F-negative patients and in some cases with idiopathic erythrocytosis (IE).1–3 These patients display a clinical phenotype and bone marrow histologic features which differ from those observed in V617F-positive PV patients. This proposed distinctive myeloproliferative syndrome would be characterized by marked erythrocytosis, absence of leukocytosis or thrombocytosis, low serum erythropoietin levels and isolated erythroid hyperplasia in the bone marrow and, in some cases, endogenous erythroid colony growth. However, information about clinical outcome of these patients and hematopoietic cell lineage involvement of these mutations is limited.

We analyzed 29 patients diagnosed with PV or IE at three institutes, who were negative for the V617F JAK2 mutation by allele specific PCR. The diagnosis of PV was centrally reviewed according to the criteria of the Polycythemia Vera Study Group.4 Idiopathic erythrocytosis was diagnosed when patients having an absolute increase in red cell mass did not fulfil PV criteria and causes of secondary erythrocytosis were excluded after a careful workup, according to the guidelines for the diagnosis, investigation and management of polycythemia/erythrocytosis.5 Nine out of 29 PV patients were positive for the V617F JAK2 mutation by quantitative allele-specific PCR and were excluded from analysis. The remaining 20 patients (11 IE and 9 PV) were the subject of the study. Eighty-six V617F-positive PV patients, diagnosed and followed at the Hematology Department of the Hospital del Mar served as the PV control group for comparison. The study was approved by the local Ethics Committee of all participating centers and informed consent was obtained according to the Declaration of Helsinki.

JAK2 mutations in the exon 12 as assessed by direct sequencing were detected in 4 (1 PV, 3 IE) out of 20 (20%) patients negative for the V617F mutation. Three different exon 12 mutations were found: K539L which was detected in 2 patients, but was found to be from different nucleotide substitutions, N542_E543del and H538_K539delinsL (Figure 1). The first two mutations had been reported by Scott et al., but the third has not yet been described.1, 2 All 3 mutations affecting exon 12 were found in heterozygosity. Confirmation of the K539L and N542_E543del was made by allele-specific PCR as previously described.1 However, a negative result was obtained for the patient carrying the K539L mutation resulting from an AAA→CTA substitution. This emphasizes that mutations in exon 12 are irregularly allocated among several nucleotides and shows how difficult it is to detect these alterations by allele-specific PCR since primers specific for each mutation are needed.

Figure 1.Sequencing chromatograms showing mutations affecting exon 12 of the JAK2 gene. The upper panels show a lysine to leucine substitution at amino acid position 539 (K539L) resulting from a AAA→TTA exchange (left) or from a AAA→CTA exchange (right). Lower panels show the deletion of amino acids 542 and 543 (N542_E543del) (left) and the deletion of amino acids 538 and 539 with the insertion of a leucine (H538_K539delinsL) (right).

To determine which blood cell populations were affected by exon 12 mutations, granulocytes, T lymphocytes, B lymphocytes, natural killer (NK) cells and monocytes were purified by cell sorting on a FACSVantage flow cytometer and cell sorter (Becton Dickinson, San José, CA, USA) with a cell purity >95% for all fractions. Platelet-rich plasma was obtained by centrifugation. In all 4 patients, mutations affecting exon 12 were detected in granulocytes and platelets. These mutations were not detected in B and T lymphocytes or NK cells, but were detected in monocytes in 2 out of 4 cases. To our knowledge, this is the first report that demonstrates a predominant myeloid lineage involvement of JAK2 exon 12 mutations.

In 7 out of 16 patients negative for both V617F and exon 12 mutations, the whole codifying sequence of the JAK2 gene was analyzed by direct sequencing but no pathogenic mutations were detected. The lack of mutations in the JAK2 gene in this subset of patients suggests the possible existence of other genetic alterations. The main clinicohematologic data at diagnosis of the 4 patients with exon 12 mutations are shown in Table 1. The 4 patients with exon 12 mutations were of female gender in contrast to 38 out of 86 (44%) V617F-positive PV patients (p=0.04). Patients with exon 12 mutations showed significantly lower platelet counts at diagnosis (231±53 vs. 601±286×10/L, p=0.001) and lower leukocyte cell counts (6.9±2.3 vs. 11.5±4.6×10/L, p=0.02) than V617F-positive PV patients, in agreement with previous reports.1, 2 No significant differences were observed in age and hemoglobin level. PRV1 was overexpressed in all 4 patients. Bone marrow biopsy at diagnosis in 3 patients showed an increased cellularity with erythroid and megakaryocytic hyperplasia.

The remaining 16 patients (9 PV, 7 IE), negative for both V617F and exon 12 JAK2 mutations, were 3 women and 13 men. There were no significant differences in age, hemoglobin level, leukocyte counts, platelet counts and serum erythropoietin levels when compared with patients carrying exon 12 mutations. When compared with the V617F-positive PV patients, no significant differences were observed concerning age, sex and leukocyte counts. However, patients negative for both V617F and exon 12 JAK2 mutations showed significantly lower platelet counts at diagnosis (308±243 vs. 601±286×10/L, p=0.001) and higher hemoglobin level (187±16 vs. 174±25 g/L, p=0.05) than V617F-positive PV patients.

Table 1.Main clinicohematologic data at diagnosis of 4 patients with JAK2 exon 12 mutations.

After a median follow-up of six years (range 5–23 years), of the 4 patients with exon 12 mutations, 2 developed major thrombotic events at 10 months and 19 years after diagnosis (Table 1). The remaining 2 patients did not present thrombotic events after a follow-up of 4.3 and 5.6 years respectively. Transformation to myelofibrosis was observed in 1 case at 20 years after PV diagnosis (patient 3). All 4 patients remain alive at last follow-up. Given these results, it seems that the clinical outcome of patients with exon 12 mutations would be similar to that reported for classical PV patients.6

In conclusion, mutations in exon 12 are detected in a variable proportion of JAK2 V617F-negative PV or IE patients and mainly affect the myeloid lineages. Although the initial hematologic data differ from V617F-positive patients, thrombotic development and myelofibrotic transformation during follow-up are similar to classical JAK2 V617F-positive PV patients.

Acknowledgments

we thank Oscar Fornas for his contribution to cell sorting experiments and Mercè Musset and Erica Torres for excellent technical assistance

Footnotes

• LM-A and CB equally contributed to this study