Anti-CD20 to Control Antibody Formation Against FVIII In Gene and Protein Replacement Therapy

Abstract 711

Neutralizing antibodies to FVIII (“inhibitors”) following delivery of exogenous clotting factor represents a major complication in the treatment of hemophilia A. Current treatment via immune tolerance induction (ITI) is effective but expensive, protracted, and imprecise. Recently, the drug rituximab—an anti-CD20 antibody used to deplete B cells—has been used with some success to treat existing inhibitors in patients. However, the potential for this drug to prevent inhibitor development has not been addressed. We used an IgG2a monoclonal anti-murine CD20 (kindly provided by Biogen Idec) to investigate the effect of B cell depletion on the immune response to human FVIII in hemophilia A mice (exon 16 deletion, C57BL6/129) and to determine the possibility of using B cell depletion to augment gene therapy. To deplete CD20⁺ B cells, mice were given two i.v. doses of 10mg/kg anti-CD20 (αCD20) two weeks apart. Depletion of peripheral blood B cells was confirmed by flow cytometry of blood, and depletion of lymph node and splenic B cells was tracked in control animals given the same regimen. Following administration of a single treatment of αCD20, peripheral blood B cells were depleted to <1% of total lymphocytes compared to 10–30% in control animals treated with isotype control antibody. One group of these mice (n=5, “αCD20+AAV8”) was given 10¹¹vg/mouse of an AAV8 vector expressing hFVIII under a liver-specific promoter (AAV8-FVIII) one week after the first αCD20 injection. Another group (n=5, “AAV8-only”) was given AAV8-FVIII but not αCD20. Ten weeks after AAV8-FVIII treatment (8 weeks after last αCD20 treatment), both AAV8-only and αCD20+AAV8 groups were “challenged” with weekly i.v. injections of hFVIII at 1 IU/mouse for 4 weeks. Animals treated with αCD20+AAV8 showed somewhat better improvements in clotting times compared with animals treated only with AAV8, albeit that no anti-FVIII was detected in either group. However, the two groups differed more dramatically in their response to subsequent FVIII challenge. Mice in the AAV8-only group developed a mean anti-FVIII IgG1 titer of 7001 (±867) ng/mL and an average Bethesda titer of 336 (±88) BU, similar to the anti-hFVIII response seen normally in mice of this strain. In contrast, mice given αCD20+AAV8 were hypo-responsive with an anti-FVIII IgG1 titer of 1609 (±868) ng/mL and 22 (±11) BU. A third group of mice (n=5, “αCD20-only”) did not receive gene transfer and were challenged similarly, starting at 4 weeks after the last αCD20 treatment, i.e. when peripheral B cells were rebounding. Two weeks later, only 1 of 5 mice had a detectable antibody response with an IgG1 titer of 513 ng/mL and a Bethesda titer of 3.6 BU. Seven weeks after the end of the initial challenge, two of these mice were again challenged with another 4-week FVIII injection cycle. One animal again had no detectable anti-FVIII, while the other had 3478 ng/mL anti-FVIII IgG1 and 126 BU, still below the normal response. Together, these data suggest that antigen exposure upon transient depletion of B cells with anti-CD20 induces significant hypo-responsiveness to hFVIII. We have now generated human CD20 transgenic hemophilia A mice to test Rituximab for this purpose.