Clinical Factors Associated With Bleeding In Ph-Negaitve MPN

Patients with Ph-negative myeloproliferative neoplasms (MPN) that include polycythemia vera (PV), essential thrombocythemia (ET) and myelofibrosis (MF), share an increased risk of thrombotic and hemorrhagic complications. Risk factors for hemorrhagic complications are less defined compared to thrombosis, but acquired von Willebrand disease (aVWD) and extreme thrombocytosis have been implicated.

To define whether aVWD, extreme thrombocytosis or treatment history associates with bleeding in Ph-negative MPN patients.

The Northwestern University Electronic Data Warehouse identified MPN patients ≥18 years, seen between 2005-2013, with available testing for aVWD and/or extreme thrombocytosis, defined by a platelet count of > 800 x 10⁹/L. Eligibility criteria also included availability of JAK2 status, complete blood count (CBC) results, clinical history and treatment history. aVWD was defined by vW antigen and/or ristocetin cofactor activity below the laboratory reference range for blood type (usually<40% for blood group O or <53% for non blood group O). We defined aVWD type by the ratio of vW antigen to activity. Associations were tested using Fisher exact test; p<0.05 was considered statistically significant.

64 MPN patients were identified: 14 with PV (22%), 35 with ET (54%), 5 with MF (8%) and 10 with MPN NOS (16%). Median age at diagnosis was 70 years (range 18-87) and 64% were women. Most patients were Caucasian (64%) with Asians, African Americans and Hispanics representing 6% each in the study population. JAK2 V617Fwas positive in 30 patients (47%). To date, 7 (11%) patients had died and one had leukemic transformation.

Bleeding complications were reported in 13 cases (20%) while thrombotic events were reported in 11 cases (17%): ACS (2), TIA (2), CVA (1), PVT (3) and DVT/PE (3). 62% of cases had neither bleeding nor thrombosis, and only 2 of 13 cases with bleeding (unrelated to warfarin) had prior history of thrombosis. Bleeding episodes included mucocutaneous sites with 4 cases of epistaxis and 3 cases of menorrhagia. There were 5 cases of minor GI bleeding not requiring transfusion (3 with lower GI and 2 with upper GI bleeding). There was 1 case of bleeding after a bone marrow biopsy. Bleeding events were more common in ET and PV than MF or MPN NOS (6, 4, 1 and 2 cases respectively). Aspirin (ASA) was discontinued in 9 of 13 cases (70%) and hydroxyurea was added or adjusted in 5 of 13 (38%) cases. The mean platelet count at the time of bleeding was 998 x 10⁹/L (range 41-1816), mean white blood count (WBC) 15.8 x 10⁹/L (range 6.4-38.6) and mean hemoglobin (Hgb) 11.5 x 10⁹/L (range 7.6-17.1). 36 patients had testing for aVWD; the median ristocetin activity was 81% (range 12-190%), vW Antigen 112% and factor VIII activity 93%. Multimer analysis was done in 2 cases. In 1 case, an absence of large multimers was reported, suggestive of type II avWD, and was normal in the other. Of 36 patients, there were 6 (17%) cases of aVWD (ristocetin activity 12%, 22%, 37%, 38%, 38% and 53% respectively); 5 with vWD type 2 and 1 with vWD type 1. Only 2 of 6 patients with aVWD had bleeding, both while on ASA (bleeding after bone marrow biopsy and menorrhagia). Using Fisher exact test, bleeding was associated with ASA therapy only (p<0.0001). Cytoreductive therapies, age, MPN subtype, JAK2 status, aVWD and CBC did not correlate with bleeding.

We identified a similar prevalence of bleeding and thrombotic events, reflecting the selective criteria of our search, since thrombocytosis better associates with bleeding and aVWD is typically evaluated in patients with bleeding or thrombocytosis. In our study, bleeding was associated with ASA intake, was often mucocutaneous in origin and typically resulted in ASA discontinuation and/or change in cytoreductives. An evidence-base supports ASA use in PV, but is lacking in other MPN. Therefore, treatment with ASA should be initiated on an individual rather than empirical basis given a potential for bleeding in some MPN patients. The mean platelet count at the time of bleeding was extremely increased, but aVWD was rare. Moreover, bleeding was infrequent in those with aVWD. Prospective studies are needed to identify the incidence and clinical relevance of aVWD in MPN, but acquired platelet dysfunction by another mechanism may be a more relevant cause of MPN-associated bleeding.

No relevant conflicts of interest to declare.