Prevalence,Type, and Risk Factors for Bleeding in a Large Cohort of Myeloproliferative Neoplasm Patients

Abstract

Background:The BCR-ABL negative myeloproliferative neoplasms (MPN), including polycythemia vera (PV), essential thrombocythemia (ET) and myelofibrosis (MF), share an increased risk of thrombotic and hemorrhagic complications. The risk of bleeding and thrombosis is less defined in MPN-Unclassifiable (MPN-U). Risk factors for hemorrhage are less defined compared to thrombosis, but acquired von Willebrand disease (aVWD) and thrombocytosis have been implicated. Because patients (pts) with CALR mutations have higher platelet counts compared to JAK2 V617F mutated pts, bleeding rates may be increased in the former group. Our aim was to define the prevalence and bleeding type and to evaluate whether aVWD, thrombocytosis, mutational status or treatment history associated with bleeding in a large cohort of MPN pts.Methods:The Northwestern University Electronic Data Warehouse identified MPN pts ≥18 years, last seen between 2005 and 2013. MPN was diagnosed based on clinical features, histopathology, and mutational status. A significant bleeding event was defined by a need for medical evaluation. Mutations in exon 9 of CALR were evaluated by PCR and fragment analysis. aVWD was defined by vW antigen (ag) and/or ristocetin cofactor activity below the laboratory reference range for blood type (<40% for blood group O or <53% for non blood group O; type defined by vW ag: activity ratio). Associations were tested using Fisher’s exact test; p<0.05 was considered statistically significant.Results I: Baseline characteristics351 MPN pts were identified: 142 with ET (40.4%), 118 with PV (33.6%), 62 with MF (17.7%), 4 with MDS/MPN (1.1%), and 25 with MPN-U (7%). Median follow-up was 5 years, median age at diagnosis was 54.7 years, and 58.1% were female. 73.3% were treated with aspirin, and 51.6% were treated with hydroxyurea. JAK2 V617F was identified in 178/288 patients (61.8%). Among 110 JAK2 V617F-negative samples, residual DNA was available in 46 and 15/46 (32.6%) were CALR mutated, including primarily type 1 (52-bp deletion; c.1092_1143del, N=11), type 2 (5-bp insertion; c.1154_1155insTTGTC, N=3) mutations, and 1 with a 69-bp deletion. CALR mutated patients included 10/29 ET (34.5%), 2/7 MF (28.6%), and 3/9 MPN-U (33%). To date, 34 patients in the cohort (9.8%) progressed to MF and 18 (5.2%) had leukemic transformation.Results II: Bleeding complications55 pts (15.6%) experienced 62 bleeding events, while thrombotic events were reported in 84 cases (23.9%). 19% of cases had both bleeding and thrombosis. In bleeding pts, the mean white blood count (WBC) at the time of diagnosis was 15.0 x 109/L, mean platelet count 803 x 109/L, and mean hemoglobin 12.6 x 109/L. Bleeding occurred at a median of 2 years after diagnosis. Gastrointestinal bleeding (GIB) was the most common type in 28/55 pts (50.9%), including 10 upper GIBs, 9 lower GIB, and 9 cases involving an unspecified location. Bleeding episodes from mucocutaneous sites were noted in 17/55 pts (30.9%; 9 with epistaxis, 1 with gum bleeding, and 7 with vaginal/uterine bleeding). There were 6 pts with intracranial hemorrhage. The mean platelet count at the time of bleeding was 473 x 109/L. Bleeding events were significantly more common in MPN-U (32%) than MF (19%), ET (10.6%), or PV (15.3%), (p=0.0163). Bleeding was associated with older age at diagnosis (59.4 yrs in bleeding pts vs 53.9 yrs in non-bleeders, p=0.03). There was no association between bleeding and mutational status (JAK2 V617F, CALR), gender, or aspirin use. The median ristocetin activity was 94% (N=38), vW ag 110% (N=37) and factor VIII activity 94% (N=28). Of 38 pts, there were 7 (18.4%) cases of VWD; 6 were acquired (5 with type 2, 1 with type 1) and 1 was congenital, respectively. Two of 7 pts with VWD (28.6%) had bleeding events.ConclusionsConsistent with prior reports, bleeding was less prevalent than thrombosis, associated with older age and, uniquely MPN-U, an entity with a less defined natural history compared to ET, PV, or MF. VWD was infrequently tested for, even less commonly identified, and less represented in bleeding cases. The role of routine testing for aVWD requires further definition. We did not find any association between bleeding and aspirin use or mutational status. However, a limited number of CALR mutated patients may have precluded recognition of any association, and future studies with a larger sample of CALR mutated patients are needed to identify its impact on bleeding. DisclosuresNo relevant conflicts of interest to declare.