The V617F JAK2 mutation is uncommon in cancers and in myeloid malignancies other than the classic myeloproliferative disorders

Four groups recently reported the existence of an activating mutation of JAK2 in many patients with one of the classic myeloproliferative disorders (MPDs).^1-4  Mutant Janus kinase 2 (JAK2) has increased kinase activity, renders BaF3 cells cytokine independent, and produces erythrocytosis in vivo. The V617F mutation is present in most patients with polycythemia vera, and a substantial proportion of patients with idiopathic myelofibrosis (IMF) or essential thrombocythemia (ET). It remains unclear whether the V617F JAK2 mutation contributes to other hematologic malignancies or to nonhematologic tumors.

Sequence analysis was used to determine the JAK2 status of 618 cell lines derived from hematologic (132 lines) and nonhematologic (486 lines representing more than 30 different tumor types) malignancies (Table 1). The V617F mutation was not detected in most of the cell lines but was present in 2 acute myeloid leukemia (AML) lines (HEL and HEL-92.1.7). Micro-satellite analysis demonstrated that these were clonally related. We therefore analyzed 211 primary hematologic malignancies (including 90 AML samples) using allele-specific polymerase chain reaction (PCR)¹  to detect the V617F allele. The mutation was present in 5 (5.6%) of the AML samples (Table 1). None of the 5 patients had the FMS-like tyrosine kinase 3-internal tandem duplication (FLT3-ITD) mutation, or the t(8;21), inv(16), or t(15;17) rearrangements. All 5 were male (P = .06, Fisher exact test) and significantly older than their unaffected counterparts (median ages at diagnosis were 68 years for V617F-positive patients and 52 years for V617F-negative patients; P = .05, Wilcoxon rank-sum test). None of these V617F-positive patients had a prior history of an overt MPD. Although a prior study did not detect JAK2 mutations in AML patients,⁵ JAK2 rearrangements occur infrequently in hematologic malignancies. The t(9;12), t(8;9), and t(9;22) translocations present in ALL and atypical chronic myeloid leukemia (CML) each disrupt JAK2, generating the translocation Ets leukemia (TEL)/JAK2, pericentriolar material 1 (PCM1)/JAK2, and breakpoint cluster region (BCR)/JAK2 proteins, respectively.^6-8  The kinase activity of the TEL/JAK2 fusion protein is required for its leukemogenic properties,⁹  consistent with the concept that constitutive JAK2 activity resulting from the V617F mutation contributes to the leukemic phenotype.

The V617F mutation was not present in patients with CMML or CLL, or in the majority of patients with BCR/ABL-positive CML or MDS. We did detect this mutation in single instances of CML and MDS (Table 1). The V617F-positive CML patient had a 12-year history of BCR/ABL-negative thrombocytosis prior to presenting with CML. His subsequent clinical course suggested the coexistence of imatinib-sensitive CML and imatinib-resistant ET.¹⁰  Similarly, the V617F-positive MDS patient had trephine features of myelodysplasia with fibrosis, consistent with an MDS/MPD overlap syndrome.

The molecular basis of the V617F-negative MPDs remains obscure. We therefore screened samples from 24 MPD patients without the V617F allele for alternative JAK2 mutations and for mutations in other components of the JAK/signal transducer and activator of transcription (STAT) pathway. All 128 coding exons of JAK1, JAK2, JAK3, TYK2, STAT5A, and STAT5B were sequenced in the peripheral blood granulocyte DNA from 19 ET and 5 IMF patients who lack the V617F allele. No mutations were detected in these individuals (data not shown).

Collectively, these data suggest that the JAK2 V617F mutation does not occur in nonhematologic cancers, that this mutation is uncommon in myeloid malignancies other than the classic BCR/ABL-negative MPDs, and that the V617F-negative MPDs are likely to reflect mutations in other molecules that modulate the JAK/STAT pathway, or mutations in different signaling pathways. We would like to acknowledge the support of Drs Jenny Craig, Robert Marcus, and Charles Crawley, and Prof Alan Warren, together with the clinical team at Addenbrooke's Hospital, Cambridge. We are grateful to Drs Anthony Bench and Wendy Erber, and the staff of the Haematological Disorders Sample Bank in Addenbrooke's Hospital for help with sample processing, and to Clare East for data management.