The Determination of the Inhibitory Activity of Some Factor VIII Type II Inhibitor Antibodies Is Biased by the Type of Factor VIII Assay.

Although type II Factor VIII (FVIII) inhibitor antibodies inhibit only partially FVIII activity, even when such antibodies are in very large excess over FVIII, they frequently cause severe bleedings in patients with acquired or congenital hemophilia A. During the preclinical evaluation of anti-FVIII antibodies as potential antithrombotic agents in baboons, we observed severe bleedings in animals treated with an antibody (mAb-LE2E9) inhibiting in vitro about 70% baboon FVIII. The administration of another antibody (mAb-LE2E9Q), inhibiting less than 20% baboon FVIII in vitro did not induce any significant hemoglobin drop by comparison to control animals. Surprisingly, mAb-LE2E9Q significantly reduced thrombus development in extracorporeal circulation models of venous and arterial thrombosis. The discrepancy between the levels of FVIII inhibition observed with these two antibodies and their in vivo activity prompted us to use modified types of FVIII assays to further evaluate in human plasma the inhibitory activity of these antibodies. When a FVIII chromogenic assay was performed with a low phospholipid concentration rather than with the standard concentration, the inhibitory activity of mAb-LE2E9 increased from 83 to 93% and that of mAb-LE2E9Q from 36 to 49%. Similarly, when activated platelets were used as a source of phospholipid, the maximal inhibitory activity of both antibodies increased to the same levels. Likewise, the use of different one stage and two stages commercial FVIII assays allowed to record maximal inhibitory activities ranging from 20 to 59% in a reference human plasma spiked with mAb-LE2E9Q. Thus, type II anti-FVIII antibodies induce qualitative FVIII alterations that are poorly detected by some FVIII assays. Variability inherent to the FVIII assays may therefore contribute to the large interlaboratory variability encountered even with a standardized method such as the Nijmegen modification of the Bethesda assay. This phenomenon may also account for the discrepancies frequently observed between severity of bleedings and residual FVIII activity in patients with type II inhibitor antibodies.