Generation of FIX variants to bypass FVIII. (A) Screening of hFIX variants for FVIII-independent clotting. Each mark represents an hFIX variant tested. hFIX variants of interest used for in vivo testing are highlighted. Values represent means of specific FIX activity calculated as FIX activity divided by FIX antigen given as a percentage. 100% is defined as full FIX activity of the WT protein in the presence of FVIII (ie, a specific activity of 200 U/mg). Plotted on the y-axis are specific activity values calculated from measurements in FIX-deficient plasma; on the x-axis are values calculated from measurements in FVIII-deficient plasma. Each dot represents the mean value from at least 6 independent tests for each variant with error bars showing SEM. (B) Illustration of amino acid substitutions for variant ITV. (C) Specific FIX- and FVIII-bypassing activity of hFIX variants (hFIX-T and hFIX-ITV) compared with hFIX-WT. (D-E) Effect of hFIX variants on clotting times by one-stage FVIII assay in the presence of inhibitory anti-FVIII Abs. Plasma samples of patients with inhibitory FVIII Abs were mixed 1:1 with FIX variant solution or buffer with a final hFIX variant concentration of 2.5 μg/mL (50% of normal FIX antigen levels) and incubated for 2 hours at 37°C to allow Ab binding. A range for normal coagulation was established using identical incubation with control plasma from healthy volunteers without inhibitory Ab–containing samples (n = 5) using the same procedure. *P < .05 and **P < .01 (ANOVA with Dunnett test for multiple comparisons with the FIX-WT group). (E) Remaining coagulation activity, inhibitor levels in Bethesda units (BU), and underlying disease (acquired hemophilia with autoimmune response against FVIII and hereditary hemophilia A with inhibitory Abs) are shown for each patient.