The necessity of repeat testing for von Willebrand disease in adult patients with mild to moderate bleeding disorders

Background: In patients with mild-to-moderate bleeding disorders (MBD), von Willebrand disease (VWD) is diagnosed at von Willebrand factor (VWF) levels <= 50 IU/dL. Although VWF levels are unstable, repeated testing for VWD diagnosis is not necessarily advised in recent guidelines. Objectives: To analyze the relevance of repeated VWF testing to diagnose VWD in patients with MBD. Methods: Data of 277 patients with MBD from the Vienna Bleeding Biobank with at least 2 separate assessments of VWF antigen (VWF:Ag) and activity (VWF:Act) were analyzed. Results: In repeated VWF measurements, 36 patients (13.0%) had "changing" VWF levels (<=/>50 IU/dL), 27 (9.7%) had persistent levels <= 50 IU/dL ("pathologic"), and 214 (77.3%) had levels >50 IU/dL ("normal"). Of the 36 changing patients, 22 (61%) were diagnosed with VWD at baseline, whereas the others only met VWD diagnostic criteria at repeated measurements. Using logistic regression, we estimated a probability of change of 26.4% (95% CI, 12.5-47.4) at baseline VWF levels of 30 IU/dL, 50.8% (95% CI, 35.6-65.8) at 50 IU/dL, 18.8% (95% CI, 12.3-27.6) at 60 IU/dL, and 1.2% (95% CI, 0.3-4.9) at 80 IU/dL. Baseline VWF was a strong predictor for changing status (Chi(2) = 49.9; P < .001), while age, sex, Vicenza score, and blood type O had limited added value (Chi(2) = 5.1; P = .278). Baseline VWF:Ag or VWF:Act cutoffs of 80 IU/dL had negative predictive values of 98.1% and 99.1% for changing status, respectively. Conclusion: Our data emphasize an overlap between patients with VWD and MBD with bleeding disorder of unknown cause and underline the need for repeated VWF testing, especially in patients with VWF levels <80 IU/dL.