B cell-activating factor modulates the factor VIII immune response in hemophilia A

Abstract

Inhibitors of factor VIII (FVIII) remain the most challenging complication of FVIII protein replacement therapy in hemophilia A (HA). Understanding the mechanisms that guide FVIII-specific B cell development could help identify therapeutic targets. The B cell-activating factor (BAFF) cytokine family is a key regulator of B cell differentiation in normal homeostasis and immune disorders. Thus, we used patient samples and mouse models to investigate the potential role of BAFF in modulating FVIII inhibitors. BAFF levels were elevated in pediatric and adult HA inhibitor patients and decreased to levels similar to those of noninhibitor controls after successful immune tolerance induction (ITI). Moreover, elevations in BAFF levels were seen in patients who failed to achieve FVIII tolerance with anti-CD20 antibody-mediated B cell depletion. In naive HA mice, prophylactic anti-BAFF antibody therapy prior to FVIII immunization prevented inhibitor formation and this tolerance was maintained despite FVIII exposure after immune reconstitution. In preimmunized HA mice, combination therapy with anti-CD20 and anti-BAFF antibodies dramatically reduced FVIII inhibitors via inhibition of FVIII-specific plasma cells. Our data suggest that BAFF may regulate the generation and maintenance of FVIII inhibitors and/or anti-FVIII B cells. Finally, anti-CD20/anti-BAFF combination therapy may be clinically useful for ITI.

Keywords: Coagulation; Cytokines; Hematology.

Figure 1. B cell cytokine levels in pediatric patients with hemophilia A.

B cell cytokines in pediatric HA patients with FVIII inhibitors (n = 24) or without FVIII inhibitors (n = 45). (A) BAFF levels. (B) APRIL levels. (C) BCMA levels via unpaired t test. Longitudinal analysis of (D) BAFF, (E) APRIL, and (F) BCMA levels in pediatric patients with inhibitors who failed immune tolerance induction (n = 4) or succeeded (n = 6) via paired t test. (G) Peripheral T-helper cytokine levels in pediatric HA patients with (red squares) and without (blue squares) inhibitors via unpaired t test. (H) Heatmap of Spearman’s correlation of Bethesda titer, α-FVIII IgG subclasses, and cytokines. Box-and-whisker plots show median with 25%–75% IQR, whiskers delineate 10th and 90th percentiles, with values outside these ranges shown as symbols. Other data plotted as mean ± SD. *P < 0.05. NS, not significant.

Figure 2. B cell cytokine levels in adult Italian hemophilia A patients.

B cell cytokine and α-FVIII IgG levels in adult HA patients with (n = 22) or without (n = 24) FVIII inhibitors. (A) BAFF levels. (B) APRIL levels. (C) BCMA levels. Receiver operating characteristics of (D) BAFF, (E) APRIL, and (F) BCMA for pediatric and adult HA patients. (G) α-FVIII IgG in adult HA patients. (H) Spearman’s correlation heatmap of B cell cytokines and α-FVIII IgG in adult HA patients. *P < 0.05, **P < 0.01, ***P < 0.001 by Mann-Whitney U test. NS, not significant.

Figure 3. BAFF levels in HA inhibitor patients treated with rituximab.

(A) Schema for rituximab therapy. Adult and pediatric HA patients with refractory inhibitors were treated with rituximab (black diamonds) and plasma samples (red drops) were obtained before and after therapy. Patients (n = 8) received concurrent FVIII ITI or not (n = 9). (B) BAFF levels before and after rituximab therapy in HA patients who did (black circles, n = 3) or did not (red circles, n = 14) achieve FVIII tolerance at the end of their ITI course. **P < 0.01 by paired t test. NS, not significant.

Figure 4. α-mBAFF antibody therapy for prevention of FVIII inhibitors in HA mice.

(A) C57BL/6-129 HA mice (n = 10–14/group) were injected with α-mBAFF antibody prior to immunization with FVIII and followed longitudinally. (B) Number of inhibitor-positive (black bars) or -negative (gray bars) mice in controls versus α-mBAFF–treated groups. (C) BAFF levels over time in the α-mBAFF (red circles) and control (black circles) groups. (D) α-FVIII IgG in the α-mBAFF group (red circles) compared to controls (black circles) on day 56. FVIII inhibitor titers (E) and α-FVIII IgG (F) after remote FVIII challenge (blue arrow) in control (black circles) and α-mBAFF–treated (red circles) mice. (G) Titers of neutralizing antibodies against AAV8. AAV8 was injected 17 weeks after mice were treated with α-mBAFF antibody and α-AAV8 antibody titers were measured before and 4 weeks after AAV8 injection (n = 6). *P < 0.05; **P < 0.01; ***P < 0.001 by Fisher’s exact (B), 2-way ANOVA (C), Mann-Whitney U (D and E), Wilcoxon’s matched signed rank (F), and paired t (G) tests. NS, not significant.

Figure 5. Combination of α-mBAFF and α-mCD20 therapy for FVIII tolerance induction.

(A) Schema for combination α-mCD20 and α-mBAFF therapy. HA-BALB/c mice with established inhibitors were treated with α-mCD20 (gray squares, n = 8), α-mBAFF (red circles, n = 6), combination therapy (blue triangles, n = 6), or no treatment (black circles, n = 5) and followed for 13 weeks. (B) BAFF levels at week 5, (C) peripheral CD19⁺ B cells at week 5, (D) inhibitor titer, and (E) α-FVIII IgG₁. *P < 0.05; **P < 0.01; ***P < 0.001 by 1-way ANOVA (B and C) or mixed-effects ANOVA (D and E). NS, not significant.

Figure 6. Combination therapy with α-mCD20 with α-mBAFF or mTACI-Fc in FVIII inhibitor mice.

(A) Schema for combination α-mCD20 and α-mBAFF or mTACI-Fc therapy. HA inhibitor mice were treated with α-mCD20 with α-mBAFF (red circles, n = 10), α-mCD20 with mTACI-Fc (purple diamonds, n = 8), or no treatment (black circles, n = 8) and followed for (B) Bethesda titer and (C) α-FVIII IgG₁. At 16 weeks from start of regimen, spleens (D) and bone marrow (E) were harvested for quantification of plasmablasts and plasma cells by flow cytometry (values normalized per million lymphocytes). (F) FVIII-specific B cell ELISPOT from splenic and bone marrow plasma cells (conducted in triplicate from n = 4 mice per group), with representative images of samples (G). *P < 0.05; **P < 0.01; ***P < 0.001 by mixed-effects ANOVA (B and C) or 1-way ANOVA (D–F). NS, not significant.