Introduction
The most serious complication in haemophilia A (HA) replacement therapy with coagulation factor VIII (FVIII) is neutralizing antibodies, i.e. inhibitors. It has been hypothesized that danger signals generated during a bleed might have an adjuvant effect on the immune response to FVIII in on-demand treatment, increasing the inhibitor risk.

Aim
To compare the antibody response to treatment with recombinant human FVIII (rhFVIII) in relation to induced knee joint bleeds and treatment without concurrent bleeds in a HA rat model.

Method
HA rats were divided into two groups: one group (n = 10) receiving three needle induced knee joint bleeds 14 days apart and a control group (n = 9) receiving three sham procedures. Three hours after each injury/sham 50 IU kg−1 rhFVIII was administrated intravenously. Subsequently, both groups continued rhFVIII treatment for another 9 weeks. Binding antibodies were analysed using an enzyme-linked immunosorbent assay and neutralizing antibodies using a Bethesda-like assay.

Results
Rats in the knee-bleed group developed a significantly faster inhibitor response and reached significantly higher inhibitor levels. In the knee-bleed group, 80% developed inhibitors vs. 33% in the control group, demonstrating a 2.4 times higher inhibitor risk when treating concurrent with bleeds.

Conclusion
FVIII treatment in relation to a bleed potentiates inhibitor development compared to FVIII treatment alone in this HA rat, indicating that bleeding is a potential danger signal. Our results support the theory that FVIII replacement therapy concurrent with a bleeding episode increases the inhibitor risk, which to the best of our knowledge, has not been confirmed in an animal model before.