Abstract
Current treatment of hemophilia A is severely limited by the development of anti-fVIII inhibitory antibodies (inhibitors). Several approaches are being pursued to overcome these limitations including the utilization of human (h) fVIII alternatives such as porcine (p) fVIII, currently in clinical trials, or bypassing agents such as recombinant human activated factor VII. During our characterization of recombinant pfVIII, we identified several species-specific differentials in biochemical and immunological properties that were predicted to impart superior therapeutic capability including greater specific activity, slower decay following activation, significantly improved biosynthetic rate, and reduced antigenicity. Some patients, however, harbor cross-reactive and inhibitory antibodies against pfVIII and thus are not responsive to pfVIII therapeutics. Recently, we characterized the biochemical properties of recombinant ovine (o) fVIII and now sought to study its antigenic properties by determining the antigenicity and residual activity of ofVIII in inhibitor patient plasmas. ELISAs measuring reactivity to B-domain deleted (BDD) hfVIII, pfVIII, and ofVIII were performed on 38 patient plasmas (28 congenital and 10 acquired patients) revealing 34 patients with reduced reactivity against pfVIII and ofVIII when compared to hfVIII. Furthermore, both orthologs displayed equivalent degrees of cross reactivity in comparison to hfVIII (P = 0.155; Mann-Whitney U Test). Investigation of the residual activity of each fVIII molecule in the inhibitor patient plasmas using a modified Bethesda assay revealed that pfVIII and ofVIII were functionally less inhibited than hfVIII with median values of 3.0, 6.6, and 27.9, respectively, but were not significantly different from each other (P < 0.05; Kruskal-Wallis One Way ANOVA). Of clinical significance, 28 patients harbored Bethesda titers above 10 Bethesda units (B.U.) against hfVIII, and of these, 16 patients displayed Bethesda titers at or below 10 B.U. against both pfVIII and ofVIII suggesting potential utility as an acute therapeutic. Finally, we investigated the relative significance of species-specific amino acid substitutions within B-cell epitopes towards conferring reduced immunoreactivity and improved activity of pfVIII and ofVIII in inhibitory patient plasma. A competition ELISA was performed using a large panel of monoclonal fVIII antibodies (mAbs) against A2 and C2 domains of fVIII, since most inhibitory Abs against hfVIII target these two domains (Markovitz et al, Blood, 2013, 121:14; Meeks et al, Blood, 2007, 110:13). Using this mAb panel, 8 A2 and 8 C2 biotinylated mAbs covering all groups of inhibitors were competed against patient inhibitor plasmas for binding to hfVIII via ELISA. Almost all A2 domain targeting mAbs classified within groups A, D, and E are highly inhibitory, while most antibodies classified into groups B and C are non-inhibitory. All group A antibodies are highly potent Type I inhibitors which inhibit >90% of fVIII activity at saturating concentrations. Ten patients were known to have A2 domain specific inhibitors as determined by homolog scanning ELISA. Four of these patients demonstrated competition with A2 mAbs. Interestingly, competition was only observed within group A and B inhibitors by two mAbs tested; Mab413, and G6. In the C2 domain, we saw significant competition in all groups except group B, with no observed specificity. Ten patients displayed competition within the C2 and of those, 7 patients retained titers above 10 B.U. against ofVIII and 6 patients against pfVIII. Performing a similar ELISA against pfVIII and ofVIII found that the same group A A2 inhibitory mAbs did not bind pfVIII and ofVIII compared to hfVIII. Collectively, these data suggest that low Bethesda titers against pfVIII and ofVIII are due to the lack of antigenic A2 domain epitopes in these molecules, while in cases where residual titers above 10 B.U. persist, it is primarily due to the presence of cross-reactive polyclonal C2 inhibitors. These data also support the utility of homolog scanning as a method to identify the most antigenic epitopes in hfVIII, and the information obtained can be utilized to rationally design less antigenic fVIII products.
No relevant conflicts of interest to declare.
Author notes
Asterisk with author names denotes non-ASH members.