Efficacy of Ruxolitinib in Chronic Eosinophilic Leukemia Associated with t(8;9)(p22;p24) and PCM1-JAK2 Fusion Gene.

Abstract 2833

The Janus kinase 2 (JAK2) gene is activated by point mutation (V617F) in patients with Philadelphia-negative myeloproliferative neoplasms (MPN), playing an important role in their pathogenesis. This has led to clinical trials on the use of JAK2 inhibitors in the treatment of MPN, and to the recent approval of ruxolitinib for treatment of primary myelofibrosis. JAK2 may also be activated by translocation and fusion with another gene. The t(8;9) (p21–23;p23–24), found in atypical myeloid and lymphoid neoplasms, fuses JAK2 with the Pericentriolar Material 1 (PCM1) gene, activating JAK2.

A 31-year-old female patient was referred to our Department because of splenomegaly (12 cm below costal margin), anemia (11.5 g/dL), leukocytosis (WBC 21.6 × 10⁹/L) with eosinophilia (eosinophils 3.5 × 10⁹/L ), and thrombocytopenia (107 × 10⁹/L ). The bone marrow biopsy was hypercellular (95%), and showed eosinophil proliferation and fibrosis. Studies of X-chromosome inactivation pattern demonstrated clonal hematopoiesis, while cytogenetic analysis revealed t(8;9)(p22;p24). There was no evidence of BCR-ABL1 fusion gene nor of PDGFRA or PDGFRB rearrangements, and a diagnosis of chronic eosinophilic leukemia, not otherwise specified (CEL, NOS) was made. The candidate genes for fusion were PCM1 on chromosome 8p22 and JAK2 on chromosome 9p24. Fluorescence in situ hybridization (FISH) on bone marrow cells with probes for PCM1 and JAK2 revealed the presence of two fused signals on der(8) and der(9), indicating the presence of a PCM1-JAK2 rearrangement. The presence of chimeric PCM1-JAK2 fusion transcript was confirmed by reverse transcription PCR (RT-PCR) in RNA from circulating granulocytes. Sanger sequencing was performed to define the fusion junctions, and this showed an in-frame fusion between PCM1 exon 36 and JAK2 exon 9. The fusion protein retains the coiled-coil domains of PCM1 and the tyrosine kinase domain of JAK2: this likely facilitates oligomerization of the PCM1-JAK2 chimera resulting in a constitutive activation of JAK2.

The clinical course of patients with PCM1-JAK2-fusion-associated neoplasms is generally poor, and allogeneic stem cell transplantation represents the only curative treatment. Unfortunately our patient did not have a compatible stem cell donor. A SNP array evaluation did not detect any additional chromosomal aberration, and the PCM1-JAK2 fusion emerged as the unique genetic lesion. We therefore considered a treatment with a JAK2 inhibitor, and more specifically a compassionate use of ruxolitinib. Following approval by the local Ethics Committee and written informed consent, in July 2011 the patient started ruxolitinib at a dose of 15 mg BID. As of July 2012, this treatment is still ongoing without any adverse effect. The patient obtained a complete clinical remission with regression of anemia, leukocytosis, eosinophilia, splenomegaly and marrow fibrosis, and with restoration of polyclonal hematopoiesis.

The cytogenetic response was assessed on bone marrow at 3, 6, and 12 months. The percentage of metaphases with t(8;9) showed a progressive decrease from baseline (80%) to the 12th month of treatment (30%). Similarly, FISH with the commercial probe ON JAK2 (9p24) Break (Kreatech Diagnostics, Amsterdam, The Netherlands), optimized to detect translocations involving JAK2 at region 9p24, revealed a progressive reduction in the proportion of rearranged nuclei. To monitor the amount of fusion transcript, we used quantitative PCR analysis for the PCM1-JAK2 rearrangement in RNA samples collected from granulocytes at 3, 6, 9 and 12 months. Quantitative PCR analysis showed an early reduction in the level of PCM1-JAK2 fusion transcript, followed by a plateau at about 20% of the baseline value.

In conclusion, the identification of the PCM1-JAK2 fusion gene in our patient with CEL provided a molecular target for treatment with the oral JAK2 inhibitor ruxolitinib, which allowed a complete clinical remission and a considerable reduction in the PCM1-JAK2 clone size. As complete hematologic remissions are unlikely in MPN patients treated with ruxolitinib, our case may suggest that ruxolitinib is more effective in patients in whom JAK2 is activated by translocation than in those in whom it is activated by point mutation. Finally, since PCM1-JAK2-fusion-associated neoplasms have a poor prognosis, clinical trials on the use of ruxolitinib should be considered in patients with these disorders.