PF669 RUXOLITINIB TREATMENT AND RISK OF B-CELL LYMPHOMAS IN MYELOPROLIFERATIVE NEOPLASMS

Abstract

Background: A recent study has shown that JAK1/2 inhibitor treatment in myelofibrosis (MF) is associated with an increased risk for aggressive B-cell lymphomas, and that detection of a pre-existing B-cell clone might identify individuals at risk. Aims: To investigate in our cohort of patients with myeloproliferative neoplasms (MPN) the risk of developing lymphoid neoplasms (LPN) and the potential association with ruxolitinib therapy. Methods: Firstly, we interrogated our MPN database to identify patients who developed a LPN during follow-up. In addition, we next screened for B-cell clonality 38 MPN patients treated with ruxolitinib (11 essential thrombocythemia ET, 13 polycythemia vera PV, 14 MF) for whom a biological sample before and after exposure to ruxolitinib was available. The analysis of immunoglobulin heavy chain gene variable (IGHV) rearrangements was performed using RNA extracted from peripheral blood mononuclear cells. The detection of a clonal IGHV rearrangement was assessed using the IGH Somatic Hypermutation Assay v2.0 kit (Invivoscribe, San Diego, CA, USA) and GeneScan analysis. Results: Within a cohort of 3069 consecutive patients with MPN (1027 PV, 1347 ET, 140 prePMF, 405 PMF, 114 secondary MF, 36 MPN not otherwise specified; median follow up 6.7 years, range 0–41), we identified 24 patients who developed LPN during follow-up. None of them had prior therapy with ruxolitinib (P >0.9); LPNs were mostly of B-cell type (including 4 chronic lymphocytic leukemia), and only 2 T-cell lymphomas were observed. Twelve out of these 24 patients were previously exposed to alkylating agents (pipobroman or busulphan) for the treatment of the myeloid malignancy. None of the 85 patients treated with ruxolitinib (16 PV, 13 ET, 56 MF) developed a LPN during follow-up (median follow-up 12.5 years, range 0–30.6). All patients were polyclonal before starting ruxolitinib; a clonal immunoglobulin rearrangement was detected only in one of 38 patients after treatment. This particular patient was affected with JAK2- and TET2-mutated post-ET MF and had a familial history of MPN. Because of resistance to hydroxyurea, in 2009 this woman was enrolled in a clinical trial with ruxolitinib; treatment was discontinued in 2012 due to progressive splenomegaly. The IGHV rearrangement (Figure 1) performed seven months after discontinuation (2012) demonstrated the occurrence of B-cell clonality, confirmed in a subsequent sample two years later (2014). A sample collected in between (2013) was discordant, may be due to the sensibility of the assay (5%). In bone marrow biopsies, either before and after therapy, a mild interstitial lymphoid component was observed (5% of the whole cellularity) and composed by T CD3+ lymphocytes with scattered small B CD20+, CD79a+ cells, interpreted as of reactive significance, not sufficient to hypothesize a diagnosis of lymphoid malignancy. A whole body computed tomography showed an asymptomatic paravertebral mass (D6-D10), interpreted as extramedullary hematopoiesis, due to its stability during follow-up (unchanged since 2010). This patient died in September 2014 due to MF in accelerated phase.Summary/Conclusion: We observed a low rate of LPN among MPN patients (24/3069, 0.78%) and did not find any associations with a previous exposure to ruxolitinib. As none of our patients showed B-cell clonality on peripheral blood before ruxolitinib treatment, our findings reinforce the view that in the absence of B cell clones, ruxolitinib treatment may be considered relatively safe and can be initiated with close monitoring.