Extra-Splenic Role of B Cell Activating Factor Blockade in Prevention of Factor VIII Inhibitors

The development of neutralizing antibodies (inhibitors) to factor VIII (FVIII) remains the most challenging complication of protein-based replacement therapy in hemophilia A (HA). Elevated levels of the B cell survival cytokine B cell activating factor (BAFF) are associated with inhibitors in HA patients and BAFF levels decrease with successful immune tolerance induction (Doshi BS et al, J Clin Invest 2021). In HA mice, localization of labeled FVIII to the spleen and subsequent depletion of splenic leukocytes has been shown to reduce and/or temporarily prevent FVIII inhibitors. In contrast, transient depletion of BAFF prevented inhibitors upfront and promoted long-term tolerance to FVIII in HA mice. Given BAFF's ability to modulate B cell pools in secondary lymphoid organs, here we compared the splenic versus extra-splenic role of BAFF in regulating the FVIII immune response in order to localize its mechanism of action.

As BAFF could regulate survival of innate-like splenic-resident marginal zone B cells, initial experiments compared depletion of BAFF or marginal zone B cells in HA mice and showed more complete inhibitor prevention with BAFF depletion (median 0 vs 16.9 Bethesda Units [BU]/ml, respectively, p < 0.01). These data supported our hypothesis that BAFF's regulatory effect is not solely reliant on its modulation of the splenic-resident marginal zone B cell compartment in HA mice. To further understand the extra-splenic role of BAFF in FVIII immunogenicity, wild-type (WT) or severe HA mice (with F8 exon 16 knockout) on the C57Bl/6 background, were treated according to three groups (n=9-11/group): 1) sham controls, 2) splenectomy alone, and 3) anti-BAFF monoclonal antibody (2mg/kg IP once) 4 days prior to splenectomy ("combined therapy"). Following splenectomy, mice were injected with recombinant full length human FVIII at 80 IU/kg weekly for six weeks. Bethesda titers were measured one week following the final immunization. In order to provide hemostasis in HA mice without FVIII exposure prior to splenectomy, mice were given liver-directed gene therapy with AAV8-mFVIIa at 5e11 vg/mouse two weeks prior to splenectomy. Circulating mFVIIa levels of 5.8 ± 2.4 ug/ml at day 10 were observed. These levels normalized hemostasis allowing successful splenectomy in 85% of animals, which is similar to our experience in WT mice.

In the combined therapy group, only a single HA mouse developed a high titer inhibitor (7.4 BU/ml) and no WT mice developed high titer inhibitors. As the presence of endogenous FVIII antigen in WT mice does not seem to significantly alter BAFF's role in preventing the alloantibody response to xenoantigen (human FVIII protein), data for WT and HA mice were combined for analysis. Although splenectomy decreased inhibitors titers compared to sham controls (23.1 vs 78.7 BU/ml), the addition of anti-BAFF antibody therapy to splenectomy significantly reduced the risk of high titer inhibitors (RR 0.15, 95% CI 0.02-0.66, p < 0.01). The median Bethesda titer in HA and WT splenectomized mice was 23.1 (0-139.4) versus 0 (0-7.4) with addition of BAFF antibody therapy. Of note, inhibitor prevention with combination splenectomy and anti-BAFF antibody therapy seen here is similar to our prior data using just anti-BAFF antibody monotherapy for FVIII inhibitor prevention in HA mice (RR 0.23, 95% CI 0.08-0.57) suggesting that splenectomy does not significantly improve FVIII inhibitor prevention in the setting of BAFF inhibition.

Our data, using a challenging HA mouse model, suggest that anti-BAFF therapy may prevent FVIII inhibitors by inhibiting BAFF's function in non-splenic lymphoid compartments. As certain B cell subtypes, including marginal zone B cells, reside in non-splenic compartments in humans, our model mimics the anticipated human immune response to FVIII. Ongoing investigations into the B and T cell compartments in lymph nodes and bone marrow will provide further insights into the location of BAFF's regulatory role in the FVIII immune response and may provide insights into potential alternative pathways in the initial FVIII immune response.