Immunization Against Influenza but Not MMR Modulates the Anti-Factor VIII Immune Response in Hemophilia Α Mice

Introduction: The etiology of inhibitory factor VIII (FVIII) antibodies in 25-30% of hemophilia A (HA) patients remains poorly understood. It is possible that concurrent exposure to inflammatory stimuli, or ‘danger signals’, with FVIII increases the risk of inhibitor formation. HA patients generally begin FVIII replacement therapy around 12 months of age, and the median age of inhibitors onset is between 15-21 months. During this time frame, patients may be exposed to vaccines, such as the mumps-measles-rubella (MMR) and seasonal influenza vaccines. Our investigation is the first to address the concern that these vaccines may serve as danger signals that augment FVIII immunogenicity.

Methods: Our studies used 8-12 week-old FVIII E17KO C57Bl6/S129 HA mice, which carry a complete knockout of all murine MHC class II molecules, and instead express a chimeric human-mouse HLA-DRB1*1501 allele associated with increased inhibitor risk. Our preliminary studies show that 40-80% of these animals develop FVIII inhibitors following treatment.

HA mice were immunized with 10x the standard human dose of live-attenuated MMR vaccine (Priorix) subcutaneously or intravenously 24 hrs prior to 4 weekly intravenous infusions of 2 IU recombinant human FVIII (rhFVIII, Advate). Mice were subsequently re-challenged with MMR 24 hrs prior to 4 biweekly infusions of 6 IU rhFVIII. Blood samples were collected retro-orbitally or via cardiac puncture. Plasma from weeks 5 and 9 was assessed for anti-FVIII IgG by ELISA; inhibitor concentrations were assessed by a Bethesda assay on week 9.

The inactivated influenza vaccine (Agriflu) was administered at standard human doses either intramuscularly or intravenously 24 hrs before, after, or concurrently with the first of 7 biweekly infusions of 6 IU rhFVIII. Week 5 plasma samples were subjected to anti-FVIII and anti-influenza IgG ELISA and Bethesda assays. Statistical comparisons were made using the Fisher’s exact Mann-Whitney U tests, as appropriate.

Results: Subcutaneous MMR vaccination exhibited no significant differences in the incidence or titres of anti-FVIII IgG compared to HBSS-injected controls at weeks 5 and 9 (n=13). Similarly, no differences in the incidence or concentration of inhibitors were detected at week 9. We next evaluated the effects of intravenous MMR immunization on FVIII immunogenicity. Surprisingly, we again observed no differences in the incidence or magnitude of the anti-FVIII immune response at weeks 5 and 9 (n=28-30).

Importantly, we found a significant decrease in the incidence of FVIII-specific IgG in mice that were immunized intramuscularly with the influenza vaccine 24 hrs after and at the same time as the first infusion of rhFVIII (t=-24 hrs: 30%, t=0 hrs: 20%, t=+24 hrs: 26% vs control: 67%; p=0.11, 0.06, 0.04; n=10-15). Similarly, there was a significant decrease in the incidence of inhibitors at all immunization time points (t=-24 hrs: 30%, t=0 hrs: 43%, t=+24 hrs: 11% vs control: 80%; p=0.03, 0.02, 0.0022). No differences in IgG titres or inhibitor concentrations were detected. When immunized intravenously with the influenza vaccine, we observed an increase in the presence of FVIII-specific IgG, but no differences in titres. However, these differences were not statistically significant and need confirmation (t=-24 hrs: 80%, t=0 hrs: 90%, t=+24 hrs 40% vs control: 50%; p=0.58, 0.14, 1.00; n=5-10).

Conclusion: These are the first experimental studies to address vaccination as a potential danger signal in the development of an anti-FVIII immune response. Our results suggest that both subcutaneous and intravenous immunization of HA mice with the MMR vaccine do not influence the incidence or magnitude of the anti-FVIII immune response. In contrast, our data suggest that intramuscular immunization with the inactivated influenza vaccine modulates the anti-FVIII immune response and may enhance tolerance induction to FVIII, possibly through antigen competition. However, this proposal awaits further confirmation. Finally, a trend in increased antibody and inhibitor incidence in intravenously immunized mice suggests that the influenza vaccine can serve as a danger signal, but is dependent on the route of administration. Our findings contradict current vaccination concerns in the treatment of young HA patients, and instead suggest an inhibitor-protective effect from influenza immunization.