Thrombotic events in myelofibrosis: relationship with clinical and morphological characteristics, JAK2V617F mutational status and autoimmune phenomena

The clinical course of myelofibrosis (MF) may be complicated by both arterial and venous (sometimes in unusual sites) thrombotic events, whose relationship with clinical risk, grade of bone marrow (BM) fibrosis, JAK2V617F mutational status and/or mutational burden, as well as with autoimmune phenomena (anti-phospholipid, anti-RBC, antiplatelets, organ and non-organ specific antibodies) is still undefined. To address this issues, we evaluated retrospectively 100 MF patients (diagnosed from 1996 to 2010, and followed at our Institute for a median of 5.7 years until September 2012). Patients, 51 males and 49 females, median age 72 years, were classified as follows: 58 primary, 32 post-thrombocythemia and 10 post-polycythemia; according to clinical risk (IPSS), 14 were LR, 60 IR-1, 14 IR-2, and 12 HR; the grade of marrow fibrosis was MF-0 in 18 cases, MF-1 in 65, MF-2 in 15, and MF-3 in 2 casesThrombotic events were observed in 25 patients, 17/25 venous events (12 deep venous of lower limbs and 5 splancnic), and 8/25 arterial thrombosis (2 acute myocardial infarction and 6 strokes); thrombosis was equally distributed among IPSS groups (31% in LR, 24% in IR-1, 27% in IR-2 and 18% in HR), and no relationship was found with primary and secondary MF; 84% of thrombotic events occurred in MF-0/MF-1 patients, even without a clear relationship with marrow fibrosis. Likewise, no association was found with Jak2 positivity (56% of cases) and allele burden (34% homozygous and 66% heterozygous), even if homozygosity was more frequent in post-PV MF (6/9, 67%) than in post-ET MF (6/15, 40%) (P=0.02). No relationship was observed between thrombosis and the presence of anti-phospholipid antibodies (found in 30% of cases), nor with other markers of autoimmunity [antiRBC (45% of cases), anti-platelets (15%), and serologic antibodies (57%)]; anti-phospholipid antibodies were mainly observed in MF-1 (57%)]; anti-phospholipid antibodies were mainly observed in MF-1 (19/30, 63%) and in IR-1 group (20/30, 67%). Given these results patient were prospectively followed for a median of 2 years, when we observed that thrombosis was associated with progression of clinical risk (p=0.026); unexpectedly, JAK2 allele burden was related to clinical risk progression (p=0.020) and leukemic evolution (p=0.041), although the short follow-up does not allow definite conclusions. Our results suggest that thrombotic events are not infrequent in MF, also in early clinical stages, and are not related with JAK2 V617F positivity nor with autoimmune markers.