The Journey of Primary Myelofibrosis to Splenic Marginal Zone Lymphoma on the Wheels of Ruxolitinib

A 55 years old woman presented with massive splenomegaly (20 cm below costal margin). Base line hemogram revealed; Hb 136 g/L, WBC 9.9 × 10⁹/L, Platelets 208 × 10⁹/L, Neutrophil 45%, Lymphocyte 36%, Monocyte 05%, myelocytes 6%, metamyelocytes 4%, eosinophil 3% and 1% blasts. Peripheral smear revealed leucoerythroblastic picture with 6 nRBCs/100 WBC and tear drop cells. Bone marrow biopsy showed mild osteosclerosis, diffuse fibrosis with megakaryocytic hyperplasia and clustering. Reticulin stain showed grade 2–3 fibrosis and Masson’s trichrome showed few collagen bundles. JAK 2 V617F mutation by ASO-PCR was positive. In view of massive splenomegaly, she was initially treated with hydroxyurea for 3 months with reduction in spleen size to 14 cm below costal margin and then she was put on ruxolitinib. CBC at that time revealed: Hb 97 g/L, WBC 9.1 × 10⁹/L Platelet 162 × 10⁹/L Neutrophil 53% Lymphocyte 35%. On follow up she was noted to have progressive increase in lymphocyte count with increasing spleen size. After 6 months of ruxolitinib her WBC count increased to 16.7 × 10⁹/L with peripheral smear showing 74% small mature lymphocytes. Peripheral blood immunophenotyping revealed 55% B cells (CD 19+) which were bright CD 20 positive and lambda restricted while negative for CD5, CD10, CD23, CD200, CD 38, CD11c, CD 25, CD 103 and CD123 (Fig. 1). This immunophenotype was consistent with splenic marginal zone lymphoma. As patient was clinically asymptomatic, it was decided to stop ruxolitinib and observe lymphocyte count before starting treatment with rituximab. She was then put back on hydroxyurea which has led to gradual normalization of lymphocyte counts (Fig. 2). At present patient is clinically well, on hydroxyurea for myelofibrosis and hasn’t developed aggressive lymphoma 24 months after stopping ruxolitinib.

Fig. 1.

Flow cytometric immunophenotyping shows ~ 60% lymphoid cells of which there are 55% B-cells, 26% T cells and 15% NK cells. These B-cells (magenta) are lambda restricted, negative for CD5, bright positive for CD20, negative for CD38, CD43, dim positive for CD200, show a variable expression of CD79b and are negative for CD23, CD123 and CD103

Fig. 2.

Serial white blood cell and absolute lymphocyte count on ruxolitinib therapy (green arrow indicates start of ruxolitinib; red arrow indicates stopping of ruxolitinib) (color figure online)

Increased risk of development of aggressive B cell lymphoma has been observed in patients with myeloproliferative neoplasm. In a large retrospective study by Porpaczy et al. lymphoma developed in 5.8% of patients of MPN treated with ruxolitinib compared to 0.36% patients treated with conventional therapies. Increased risk was seen in patients with primary myelofibrosis (9.7%). Of note all the patients who developed lymphoma had a preexisting B cell clone in the bone marrow which was seen in 16.8% of all patients with myelofibrosis [1]. However, another study from Italy reported by Rumi et al. did not find increased risk of lymphoma in MPN patients treated with rituximab. Moreover, they did not find preexisting B cell clone in peripheral blood of MPN patients [2]. Development of indolent B cell lymphoproliferation has not been reported in literature. Our patient developed a low-grade B cell lymphoproliferation resembling splenic marginal zone lymphoma which regressed on withdrawal of ruxolitinib. We would like to mention that our patient might had a preexisting B cell clone as there was mild peripheral blood lymphocytosis to the tune of 3.2–3.5 × 10⁹/L before starting ruxolitinib.

We conclude that patients on ruxolitinib should be followed for development of lymphoma. This should include careful evaluation of splenomegaly and peripheral blood smear for lymphocytosis. Lymphoma arising in such a setting may be indolent and not always aggressive.