Common VWF Haplotypes in Normal African-Americans and Caucasians Recruited into the ZPMCB-VWD and Their Impact on VWF Laboratory Testing.

Abstract

Von Willebrand disease (VWD) is a common bleeding disorder that has been reported to affect up to 1% of the population. Diagnostic testing for VWD relies on specific tests of von Willebrand factor (VWF) that include VWF antigen (Ag) and VWF ristocetin cofactor activity (RCo). Variability in these tests, especially in the RCo, has the potential to affect diagnosis of VWD. Clinically, the RCo/Ag ratio is used to identify patients with type 2M VWD, of which 35% of our local type 2M index cases were of African-American descent. As part of the ZPMCB-VWD, a large study of both healthy controls and patients with VWD, we evaluated VWF Ag, RCo, multimers, EU bleeding score, and VWF gene sequencing to look for mutations and/or common polymorphisms (SNPs) that might contribute to RCo measurement. Healthy controls completed a computerized version of the EU bleeding score and provided blood for clinical VWF testing. Since platelet VWF binding primarily involves the A1 domain of VWF, exon 28 gene sequencing was analyzed and common SNPs were identified, particularly in African-Americans (AA) with altered RCo/Ag ratios. Statistical comparisons were performed using t-tests. For the AA control group, the presence of specific exon 28 SNPs, including I1380V, N1435S, and D1472H, correlated with a low RCo/Ag. In controls, the 3 SNPs occurred together in 22% of AA and 1.5% of Caucasians. For AA controls with all 3 SNPs, the mean Ag was 155, RCo 115, and RCo/Ag 0.77, while for AA without the 3 SNPs, the mean Ag was 129, RCo 129, and RCo/Ag 1.01. The difference in RCo/Ag ratio was significant (p<0.001). In comparison, the Caucasian controls without the 3 SNPs had a mean Ag of 108, RCo of 115, and RCo/Ag of 1.08. When only D1472H was considered, a significant difference in RCo/Ag ratio was noted for both African-American and Caucasian controls with D1472H (present in 63% of AA and 24% of Caucasian controls). For AA controls with D1472H, the mean RCo/Ag ratio was 0.82 (range 0.42–1.16) compared to 1.01 for those without the SNP (p<0.001). The Caucasian controls with D1472H had a mean ratio of 0.87 (range 0.57–1.17) compared to 1.08 for those without the SNP (p<0.001). EU Bleeding Score was not significantly affected by either race or SNP status. All Caucasian controls with D1472H were heterozygous, while 14% of the African-American controls were homozygous. The mean RCo/Ag for the homozygous controls was 0.71, compared to 0.86 for the AA heterozygous controls (p<0.025). In order to determine if the exon 28 SNPs intrinsically altered the measurement of VWF, the 3 SNPs were engineered into recombinant VWF and expressed in HEK293T cells. VWF Ag and RCo were performed on the purified expressed VWF. The RCo/Ag ratios for the recombinant expressed VWF were decreased for 1380V and 1472H, while the construct containing all 3 SNPs showed an even lower ratio (59% of wild-type RCo/Ag ratio). These data suggest that specific exon 28 SNPs may contribute to low RCo/Ag ratios, especially in the African-American population. Since the RCo assay involves ristocetin binding to VWF, mutations (and polymorphisms) in VWF may affect the measurement of “VWF activity” by this assay and might not reflect true hemorrhagic risk.