Background

Bone marrow (BM) fibrosis is commonly recognized in the setting of primary myelofibrosis (PMF) and other myeloproliferative neoplasms. However, a variety of non-malignant conditions are associated with myelofibrosis (MF). While PMF is associated with a limited survival, other causes of MF such as AIMF appear to have a favorable clinical course (Pullarkat et al, Am J Hematol 2001;116:211). We conducted a retrospective chart review of patients diagnosed with non-malignant MF to investigate clinical presentations, therapies, and outcomes of AIMF.

Methods

Patients diagnosed with MF were identified from pathology records at Los Angeles County-USC Medical Center from 1999-2013. Histories were reviewed for the clinical indication for BM biopsy, medical conditions known to be associated with MF, and clinical course. Patients were excluded if they met diagnostic criteria for PMF or other neoplastic hematologic disorders associated with MF. BM specimens were assessed as described by Pullarkat et al: (1) Presence of diffuse reticulin fibrosis (European Consensus Score; Thiele et al, Haematologica, 2005;90:1128).; (2) Lack of clustered or atypical megakaryocytes; (3) Lack of myeloid or erythroid dysplasia, eosinophilia, or basophilia; (4) Lymphocyte infiltration; (5) Lack of osteosclerosis; (6) Absent or mild splenomegaly; and (7) Presence of autoantibodies.

Results

Thirteen patients with MF were identified (11 F, 2 M; 23-73 y.). The mean follow up was 3.7 years. 12 presented with both anemia and thrombocytopenia, and represented the most common indication for BM biopsy. 1 of 12 also had severe leukopenia. The remaining patient presented with thrombocytosis. Clinical presentation included symptoms of fatigue, fever, lightheadedness, chest pain, and dyspnea. The spleen was enlarged in 1 patient, borderline enlarged in 4, normal in 6, and unknown in 1. 1 patient had a history of splenectomy. All patients had one or more autoantibodies. These included: ANA (8 patients), dsDNA (1), RF (4), CCP (2), DAT (7), lupus anticoagulant (3), anticardiolipin (2), atypical pANCA (1), anti-MPO (1), anti-SMA (3), anti-SSA (2), and anti-SSB (2). Medical co-morbidities included systemic lupus erythematosus (2 patients), rheumatoid arthritis (2), autoimmune hepatitis and primary sclerosing cholangitis (1), dermatomyositis (1), antiphospholipid antibody syndrome (1), and autoimmune hemolytic anemia (AIHA) (7). 3 AIHA patients had Evan’s syndrome.

BM cellularity ranged from normocellular to markedly hypercellular. The reticulin stain showed mild to moderate increase in fibrosis (MF1-2/3). Megakaryocytes were increased in 8 patients, normal in 3 patients, and decreased in 2 patients. Megakaryocytes were normal in appearance and distribution. No myeloid or erythroid dysplasia, eosinophilia, or basophilia were noted. All cases demonstrated increased small lymphocyte infiltrates. IgG4 immunostaining was negative in 4 of 4 specimens tested.

In 1 patient with Evan’s syndrome, MF was attributed to chronic romiplostim administration. 1 patient with thrombocytosis was treated as JAK2-negative essential thrombocythemia with normalization of his platelet count after interferon therapy. 4 of 11 remaining patients had a resolution in their cytopenias with steroid treatment with or without additional immunosuppressive agents. Resolution of MF was seen on post-steroid therapy BM biopsy in 2 of these patients. 3 had partial response in their cytopenias when treated with similar immunosuppressive regimens. Thrombocytopenia improved in 1 patient after splenectomy done for an unclear indication. 3 patients died due to unrelated sepsis.

Conclusion

AIMF is an under recognized cause of MF in patients presenting with cytopenias that responds to steroids and has a good clinical outcome in the majority of patients. While the spectrum of clinical presentations is broad, the BM findings are similar between cases and different from those of PMF. The prognostic and therapeutic differences between AIMF and PMF make distinction between these entities imperative for the treating physician.

Disclosures:

No relevant conflicts of interest to declare.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution