Discrepant low von Willebrand factor activity results on the ACL TOP analyzer are frequent in unselected patients with myeloproliferative neoplasms and show no correlation with high-molecular-weight multimer loss or bleeding phenotype.

BACKGROUND

Bleeding complications are common in patients with myeloproliferative neoplasms (MPNs), with a subset developing acquired von Willebrand disease. Despite this association, a wide spectrum of von Willebrand factor (VWF) abnormalities are described, and the performance of modern assays remains unclear.

OBJECTIVES

To comprehensively describe the pattern of VWF laboratory abnormalities in the MPN population.

METHODS

We collected samples from 74 unselected clinic patients with MPNs to evaluate VWF quantitatively and qualitatively via multiple methods, correlating findings with a retrospective analysis of clinical bleeding data. VWF assays were performed on both ACL TOP (Instrumentation Laboratory) and Acustar (Instrumentation Laboratory) analyzers using HemosIL reagents (Instrumentation Laboratory), along with multimer analysis by gel electrophoresis.

RESULTS

Functional VWF measurements were not concordant between assays, with a median ACL TOP VWF glycoprotein IbR to antigen ratio (VWF:GPIbR/VWF:Ag) of 0.57 (IQR, 0.43-0.71) compared to a median Acustar VWF:GPIbR/VWF:Ag of 0.91 (IQR: 0.82-1.03;P < .001). The ACL TOP showed disproportionately lower results, with 73% of patients having a ratio <0.7. Despite this, no patient experienced loss of high-molecular-weight multimers by gel electrophoresis. An inverse relationship was observed between platelet count and functional ratios on both ACL TOP (R2 = 0.20; P < .001) and Acustar (R2 = 0.18; P = .0011) analyzers. While clinically significant bleeding events were relatively common (11% patients), there was no association with VWF assay abnormalities, and generally, an alternate cause(s) was identified.

CONCLUSION

Discrepancies in functional VWF assays are common in patients with MPN, particularly by ACL TOP VWF:GPIbR. Based on our limited series, a VWF functional to an antigenic ratio of <0.7 ("type 2 pattern") alone is poorly predictive of bleeding risk.