Role of Von Willebrand Factor in Female Carriers from an Extended Family with Mild Hemophilia A

Background: Hemophilia A (HA) is an X-linked disorder that primarily affects males, although female carriers can exhibit bleeding phenotypes. Factor VIII activity levels (FVIII:C) in XX females are influenced by X chromosome inactivation (XCI), a process that silences each parental X in a proportion of cells. XCI skewing can decrease FVIII expression by preferentially inactivating the normal X chromosome. FVIII:C is further modulated by factors such as ABO blood groups and von Willebrand factor (VWF). The D'/D3 domain of VWF binds circulating FVIII protein, preventing proteolytic degradation. In type 2 Normandy von Willebrand disease (2N VWD), D'/D3 mutations decrease affinity to FVIII and result in bleeding events similar to mild/moderate HA. Variants in VWF also affect the pharmacokinetics of recombinant FVIII. Current clinical screening tests detect the 3 VWF mutations responsible for >90% of 2N, but report variants not directly responsible for 2N as clinically benign. However, common polymorphisms are known to affect FVIII:C in normal individuals. Therefore, a better understanding of how specific alterations in VWF modulate HA phenotype is necessary to interpret clinical presentation and refine management with factor concentrates. This is particularly important in HA individuals carrying mild/moderate mutations. To evaluate VWF variants in HA carriers, we focused on an extended pedigree that includes four obligate carriers from a family with mild/moderate HA (FVIII: p.Ala723Thr). FVIII:C varied and largely correlated with XCI skewing. Nevertheless, the FVIII:VWF interaction prompted us to identify VWF gene variants that could further modulate FVIII:C and contribute to bleeding in this family.

Methods: To identify D'/D3 VWF variants that impact FVIII binding, primers were designed to amplify exons 17-20 and 24-27 on chromosome 12. A chromosome 22 pseudogene, with 97% identity to VWF exons 24-34, complicated primer design. Primers specific to VWF were selected by targeting regions that differed from the pseudogene and were verified by digestion with a restriction enzyme unique to each chromosome 12 exon. The PCR products were amplified and sequenced from the four female carriers and a control male relative.

Results: After excluding the three most common mutations responsible for 2N, seven other variants were identified. Four of these were intronic polymorphisms and a synonymous variant at p.Asn1138 not associated with VWD and presumed to be clinically benign. All but one of these have been described in normal individuals. Two females were heterozygous for missense variant rs1063856 (p.Thr789Ala) and synonymous polymorphism rs1063857 (p.Tyr795=) that are in linkage disequilibrium and are likely to impact FVIII:C and VWF antigen (VWF:Ag) levels. These common variants, found in up to 36% of Caucasians and 58% of African Americans, are reported to increase VWF:Ag and FVIII:C jn heterozygotes (9 IU/dL and 7 IU/dL respectively). Neither ABO blood groups nor XCI skewing could fully explain the differences in FVIII:C activity observed with this variant.

Conclusions: We propose that VWF variants rs1063856/rs1063857 may contribute to FVIII:C differences between two females in the pedigree with similar XCI skewing. We conclude that consideration of VWF variants is important for fully understanding bleeding phenotype and treatment responses in female carriers and males in families with mild/moderate HA. These findings support the need for expanded studies into the role of FVIII and VWF variant interactions in additional unrelated HA individuals.