Background: Hemophilia A (HA) is an X-linked recessive disorder that affects males, whereas female carriers are presumed asymptomatic if Factor VIII activity levels (FVIII:C) fall within normal range. However, FVIII:C does not always correlate with bleeding phenotype, leading to an underappreciation of bleeding sequelae in females. Therefore, it is clinically important to identify HA carriers at higher bleeding risk. FVIII expression in HA carriers is influenced by X chromosome inactivation (XCI), a process that silences one X in XX females such that each cell has a random probability of inactivating either parental X. However, rare female carriers of X-linked disorders can be severely affected if XCI is skewed and the mutant X is preferentially active. How XCI skewing modulates bleeding in mild/moderate HA is less well understood. HA bleeding may be also affected by variants in factors influencing FVIII binding and clearance, including Von Willebrand Factor (VWF) and ABO blood type. To better understand HA carrier bleeding tendency, we analyzed a family that segregates a mild/moderate HA mutation (F8: c.2167G>A). Four carriers in this pedigree have FVIII:C that approach affected males, necessitating prophylaxis prior to surgical procedures. We hypothesized that bleeding in these carriers can be largely explained by XCI skewing, but additional variants may fine tune FVIII:C.

Methodology: FVIII levels were assessed by one stage (FVIII:C) and chromogenic (FVIII:CR) assays. At least two plasma samples spanning >3 years from each female were tested in duplicate with each FVIII assay. To address XCI skewing, we performed methylation-based assays at the Androgen Receptor (AR) and Fragile X Mental Retardation 1 (FMR1) loci. At least three independent PBMC DNA samples from each female were evaluated. Additionally, we screened VWF regions known to influence FVIII:C (exons 18-20, 24-27).

Results: For each female, results between XCI assays were indistinguishable (r2 = 0.99). Two of four females had pronounced skewing (≥80:20); a third had measurable skewing (67:33). Importantly, all three predominantly expressed the mutant paternal allele. Familial XCI skewing argues for a genetic cause. However, XIST, the major regulator of XCI, lacked promoter alterations. Importantly, there was linear correlation between XCI skewing and FVIII:C measured by FVIII:C or FVIII:CR assays (r2 = 0.77 and 0.83, respectively). One female with random XCI, had FVIII:C considered hemostatic (median 51%, range 43-67), whereas the other females with skewed XCI had levels <40% (16%, range 14-20, 28%, range 26-32, and 30%, range 23-35). Notably, two females had similar FVIII:C (30% and 28%) but a greater XCI skewing discrepancy (80:20 vs. 67:33). While these two females were heterozygotes for VWF p.Thr789Ala, reported to be associated with 7% higher FVIII:C, neither ABO blood type nor any additional VWF variants known to affect FVIII binding differentiated these two individuals. Therefore, it is likely that XCI skewing primarily explains their bleeding tendency.

Conclusions: Our results indicate that HA carrier bleeding phenotypes are multifaceted, and the major determinant of FVIII:C is XCI. These results also suggest that even moderate XCI skewing could influence clinical bleeding in HA carriers. However, random XCI in one female explains FVIII:C but does not fully negate bleeding tendency, emphasizing the complexity of carrier phenotype. These findings provide justification for an expanded study of carriers in unrelated pedigrees using a comprehensive approach to include XCI assays.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.