Screening for Acquired Von Willebrand Syndrome in Myelofibrosis - Poor Correlation with Bleeding Risk

Background: Thrombosis and bleeding are underestimated in myelofibrosis (MF), but those events are associated with increased morbidity and mortality. Acquired von Willebrand syndrome (AvWS) is a major concern at the diagnosis of myeloproliferative disorders and might be associated with significant bleeding events. Although this condition is expected to be related to extreme thrombocytosis, it is also present in patients with lower blood counts. Underlying mechanisms leading to AvWS are not still fully understood, and its prevalence in MF is unknown.

Aims: We evaluated a cohort of MF patients (pts) to assess the prevalence of AvWS and its correlation with bleeding events.

Methods: This is a cross-sectional study encompassing MF adult pts followed at Hospital das Clinicas da Faculdade de Medicina da Universidade de São Paulo, Brazil. Screening for AvWS was performed with von Willebrand factor antigen (vWF:Ag) and activity levels using the ristocetin cofactor assay (RCO:vWF). AvWS was defined as vWF:RCO/vWF:Ag ratio < 0.7. In addition, CD40-ligand was measured to study platelet activation. All patients signed informed consent forms.

Results: Samples from forty-six patients (52% male) were collected for this study. MF was primary in 33 cases (72%) and post-ET/PV in 13 cases (28%). Median age was 70 years (IQR 63-80), and the median time from diagnosis to sample collection was 5.3 years (2.8-9.2). JAK2 V617F was the most frequent driver mutation (46%), followed by calreticulin, and MPN W515K (33% and 7%, respectively). Sixteen patients (35%) had a DIPSS score intermediate-2 or high. None of the patients reported signs or symptoms of bleeding.

Median level of von Willebrand factor antigen (vWF:Ag) and activity of vW (vWF) were 148% (IQR 125-169) and 71% (IQR 46-100), respectively. Median vWF:RCO/vWF:Ag was 0.48 (IQR 0.32-0.62), being <0.7 in 35/46 (76%) samples. Patients with intermediate-2 and high DIPSS presented a greater median level of vWF:Ag compared with patients with lower DIPSS (189% vs. 142%. vWF/vWF:Ag ratio did not correlate with platelet counts. Concomitant use of aspirin did not influence vWF antigens or activity levels. The median plasmatic levels of CD40 ligand were 5.6 ng/mL (IQR 2-14). We did not observe increased levels of CD40L in patients with higher levels of platelets or according to DIPSS classification.

After collecting the samples, patients were followed for 30 months (range 1-44). Seventeen patients died after the study, mostly related to disease complications. In addition, one major bleeding event was reported, and one patient had a stroke.

Conclusion: We found a high prevalence of aVWS defined by the vWF:RCO/vWF:Ag ratio, even in the absence of thrombocytosis. Mechanisms of this qualitative change in the homeostasis of the Von Willebrand Factor remain elusive. One possible explanation is consumption by an increased platelet activation, corroborated by the high level of sCD40l found. The presence of laboratory aVWS criteria did not associate with clinically significant bleeding, and a diagnosis of aVWS based only on laboratory criteria should be made with caution in patients with MF. Different screening methods for this population are warranted. Additionally, the best approach of bleeding and surgery has yet to be determined.

No relevant conflicts of interest to declare.