Differentiation of Patients with Symptomatic Low von Willebrand Factor from Those with Asymptomatic Low von Willebrand Factor

Authors: Grabowski EF, Van Cott EM, Bornikova L, Boyle DC, and Silva RL.

Publication: Thromb Haemost; 120,5:793-804. May 2020

Affiliations: Department of Pediatric Hematology/Oncology, Massachusetts General Hospital for Children, Massachusetts General Hospital, and Harvard Medical School, Boston, Massachusetts, United States; Department of Pathology, Special Coagulation Laboratory, Massachusetts General Hospital, and Harvard Medical School, Boston, Massachusetts, United States; Department of Medical Hematology/Oncology, Massachusetts General Hospital, and Harvard Medical School, Boston, Massachusetts, United States.

Abstract: BACKGROUND: Accurate diagnosis of symptomatic low von Willebrand factor (VWF) remains a major challenge in von Willebrand disease (VWD). However, present tests do not adequately take into account flow forces that, at very high shear rates, reveal a weakness in the VWF-platelet glycoprotein glycoprotein Ib bond in normal subjects. The degree of this weakness is greater in symptomatic, but not asymptomatic, low VWF. OBJECTIVE: The aim of this study is to distinguish patients with symptomatic low VWF (levels in the 30-50 IU/dL range) from those with asymptomatic low VWF and normal subjects. METHODS: We measured platelet adhesion (PA)/aggregation in our novel microfluidic flow system that permits real-time assessment of PA (surface coverage) and PA/aggregation (V, aggregate volume) using epifluorescence digital videomicroscopy in flowing noncitrated whole blood at 4,000 second(-1). Blood samples from 24 low VWF patients and 15 normal subjects were collected into plastic tubes containing 4 U/mL enoxaparin. MetaMorph software was used to quantify rates of PA and V increase. RESULTS: Rates of PA increase showed a bimodal distribution, with values for 16/24 patients (Group I) all below the 2.5th percentile of normal, and values for 8/24 patients (Group II) similar to controls. Bleeding scores (mean +/- standard error) were 5.50 +/- 0.45 versus 2.75 +/- 0.45 (p = 0.00077), and 10 clinically significant bleeding events were observed in seven versus zero (p = 0.0295) Group I and Group II subjects, respectively. CONCLUSION: The present approach may offer a definitive means to distinguish symptomatic low VWF from either asymptomatic low VWF or normal controls.