Immunogenetics of factor VIII inhibitor development

Objective: Progressive destruction of joints resulting from recurrent intra-articular haemorrhage represents the major morbidity resulting from haemophilia A or B. In addition to systemic clotting factor replacement, therapies localized to haemophilic joints may provide adjunctive protection. In a factor VIII^-/- mice model, we investigated if extra-vascular delivery of recombinant human clotting factor VIII (rhFVIII) via intra-articular (IA) injection can prevent bleeding-induced joint damage, and also examined the possibility that IA delivery of FVIII carries greater risk of developing anti-rhFVIII inhibitor antibody. Methods: FVIII^-/- mice received rhFVIII by inserting a 30.5 G needle into the left knee joint, along with a range doses of FVIII(100, 25 and 5 IU kg⁻¹) in 5 μL, normal saline as the control. Comparison group received the same needle injury and intravenous (IV) rhFVIII (100, 25 and 5 IU kg⁻¹). 14 days after injury, both knee joints were collected for histological examination. To exclude the possibility that IA clotting factor was entering into circulation, mice received 100 IU kg⁻¹ rhFVIII IA, and FVIII activity was measured by aPTT. To see if IA rhFVIII delivery can carry greater risk of developing anti-FVIII antibody, mice were treated with a total dose of 300 IU kg⁻¹ rhFVIII over 10 days, either by IA or IV. 14 days after exposure, anti-FVIII was detected. After induction of anti-FVIII antibody by IV rhFVIII, mice were subjected either to needle puncture under coverage of bypassing agent (FEIBA) 100 IU kg⁻¹ or 100 IU kg⁻¹ IV rhFVIII, or needle puncture with 25 IU kg⁻¹ rhFVIII. Control mice received needle puncture with normal saline. Two weeks later, knee joints were collected for histological examination. Summary: Mice receiving only saline at the time of needle puncture developed synovitis (mean score 5.0 ± 0.5). Mice treated with 25 IU kg⁻¹ IA rhFVIII developed better protection than mice treated with 100 IU kg⁻¹ IV rhFVIII (lower pathology score for IA, 0.733 ± 0.278 vs. IV 2.57 ± 1.70) and even better protection was achieved by the dose of 100IU IU kg⁻¹ IA (Pathology score of 0.25 ± 0.31). IA injection of 100 IU kg⁻¹ rhFVIII did not lead to increased circulating FVIII activity at any time point up to 48 h. In IV-treated mice, 100% of mice developed anti-FVIII antibody (8.06BU), while only 50% of mice developed anti-FVIII inhibitor at the lowest detection limit (0.61BU). In the presence of inhibitory antibody, only 46% of mice receiving IV FVIII survived the needle injury, 58% with FEIBA and 100% of mice survived with 25 IU kg⁻¹ FVIII IA injection. In the saline-injected control mice, needle injury led to a mean pathology score of 6.8. Neither IV FVIII nor FEIBA provided effective protection, with pathology scores of 6.3 and 5.4, respectively. Surprisingly, 25 IU kg⁻¹ IA rhFVIII produced a pathology score of only 1.7. Conclusion: Extravascular rhFVIII in the joint space can contribute protection against bleeding-induced joint damage. Intra-articular rhFVIII delivery did not induce greater risk of inhibitory antibody formation in FVIII knockout mice than circulating factor VIIII challenge; in fact, a lower incidence was observed. In the presence of anti-FVIII inhibitory antibodies, IA delivery of FVIII still can offer protection from bleeding-induced joint damage.