Introduction: Hemophilia A (HA) occurs secondary to >1000 variants in the F8 gene resulting in decreased/absent production of factor (F) VIII. FVIII:C (coagulant) levels in persons with mild/moderate HA are most impacted by their genotype. ABO blood type (BT) may also impact endogenous FVIII:C levels, but this impact is difficult to assess in a genetically heterogenous group of HA subjects.

Canada has a population of mild-moderate HA secondary to a unique large single genotype: c.6104T>C, p.Val2035Ala. This variant arose from a founder effect that occurred about 200 years ago in Twillingate, Newfoundland. We hypothesized that subjects with the Twillingate genotype (despite having the same genotype and thus producing similar amounts of FVIII) would show different endogenous FVIII:C levels related to their BT, as BT impacts FVIII (and VWF) clearance. Our study objectives were to understand the impact of BT on endogenous FVIII:C levels whilst controlling for genotype, and to explore the impact of BT on DDAVP responsiveness.

Methods: This was a retrospective, single-center cohort study. Males with a confirmed Twillingate genotype treated at our center between 1980-01-01 to 2022-03-01 were eligible. For each subject, ABO BT, VWF antigen (Ag) and activity, and FVIII:C levels were determined by chart review. We recorded the lowest FVIII:C baseline level; for subjects who had >1 FVIII:C level (done in a non-bleeding state) we also determined their mean baseline FVIII:C level. For DDAVP challenges, the age at the challenge, dose, and route of administration were recorded, as well as pre and 1-hr post FVIII:C levels. Based on 1-hr post-DDAVP FVIII:C levels (IU/mL), subjects were classified on their DDAVP response as responders (≥0.50), partial responders (0.30 to 0.49), or non-responders (<0.30). Median, IQR, mean, and SD were calculated for baseline FVIII:C and VWF:Ag levels, stratified by age (<12; ≥12 y) and by BT (O; non-O). T-tests and Mann-Whitney U (MWU) tests were used to compare normally and non-normally distributed data, respectively. The Fisher-Freeman-Halton Exact Test was used to compare responses to DDAVP between O and non-O subjects. Statistical significance was considered as p≤0.05. All analyses were performed using IBM SPSS Statistics Version 28.

Results: 20 males were included: 9 were O BT and 11 were non-O BT (10 were A, 1 was B). There were statistically significant differences between the 2 ABO groups in their median lowest FVIII:C levels: 0.05 IU/mL (O BT) vs. 0.08 IU/mL (non-O BT); p=0.05 (MWU) and in mean FVIII:C levels: medians: 0.08 IU/mL (O BT) vs. 0.12 IU/mL (non-O BT); p=0.02 (t-test). Too few subjects had FVIII:C levels reported at ≥12 y of age to perform statistical analyses.

Fifteen subjects (8 O BT: 7 non-O BT) underwent a DDAVP challenge; 13 performed at our center and 2 elsewhere. Median 1-hr post-DDAVP FVIII:C levels were 0.32 IU/mL (O BT) and 0.43 IU/mL (non-O BT); p=0.04 (t-test).

The absolute FVIII:C increase with DDAVP was not statistically different between the 2 groups: median increase of 0.22 IU/mL (O BT) vs. 0.28 IU/mL (non-O BT); p=0.10 (t-test) nor was the FVIII:C fold increase: median increase 3.3-fold (O BT) vs. 3.3-fold (non-O BT); p=0.51 (t-test). Among O BT subjects, none were responders, 5 were partial responders and 3 were non-responders. In the non-O BT subjects, 2 were responders and 5 were partial responders. The variation in response between the 2 groups failed to achieve statistical significance (p=0.14).

Given uncertainties in DDAVP challenges (e.g. dose, route of administration) done elsewhere, we also analyzed the DDAVP challenge results just for the 13 subjects from our center. The median 1-hr post-DDAVP FVIII:C levels were 0.30 IU/mL in 7 O BT subjects and 0.45 IU/mL in 6 non-O BT subjects; p=0.02 (t-test).

Conclusions: This study shows that in HA subjects with an identical genotype, BT significantly impacts FVIII:C levels. Although only 1 unique genotype was studied, we believe that other hemophilic variants leading to mild HA will behave similarly. This may explain why in subjects with mild HA, there is an overrepresentation of O BT. Persons with non-O BT and HA variants that only slightly reduce FVIII production may have FVIII:C levels that are not as reduced and thus may never present with bleeding and be diagnosed with HA. In contrast, if such persons have O BT, they are likely to show much lower FVIII:C levels and thus more likely to bleed and be diagnosed with HA.

Blanchette:Novo Nordisk: Research Funding; GC Biopharma: Research Funding; Grifols: Research Funding.

Author notes

This icon denotes a clinically relevant abstract

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution