9%. In comparison, the prevalence of NNA in previously published works using ELISA assay has varied between 12.2% and 53.8% [1, 3, 7, 14], and in studies using fluorescence-based immunoassays R788 (x-MAP), a prevalence of 18.1% to 45.4% [12, 13, 15] has been reported. Several factors could explain this large variation: sample size, differences in patient characteristics (disease severity, age, mutation) and experimental design, thus, comparison

of prevalence with previously studied cohorts should be done with caution. In addition, our combined study cohort includes brother pairs, which could bias the outcome. Furthermore, an objective of HIGS is to study families in which one or more member has an inhibitor, yielding a higher prevalence of inhibitors in the overall patient population. Importantly, using cohorts of sibling pairs, we have, for the first time, been able to evaluate and compare the entire antibody response within families. Our data clearly show that the interpretation of the immune response towards the deficient factor in terms of antibody formation will AZD0530 nmr be clearly influenced by the use of an immunoassay like ELISA, instead of just capturing the immune response

by the Bethesda assay. In addition, in our cohort, the proportion of families in which all siblings developed antibodies, either of the inhibitory or non-neutralizing type, was relatively higher, supporting the concept of a genetic predisposition for the immune response to occur. When characterizing the non-neutralizing antibodies, we found that their specificity was not consistent against all three products tested in our assays. It has been suggested that NNA towards FVIII are exclusively directed towards epitopes in the non-functional B-domain of find more the FVIII protein [3, 14]. Thus, plasma with such antibodies should test positive to FL-rFVIII preparations,

but negative to the BDD-rFVIII. In agreement with this, 13 of our plasma samples (56.5%) were positive towards FL-rFVIII, but not towards BDD-rFVIII (Table 1). However, in the remaining 10 samples, NNA were found positive towards the BDD-rFVIII product, indicating antibody reactivity towards the functional domains as well. Unfortunately, all products used for treatment in the patients were not known, but the findings are in agreement with those of Lebreton et al. who recently showed that 73.7% of NNA were restricted towards epitopes of the heavy chain (A1-, A2- and B-domains) and 18.4% were directed towards the B-domain [13]. Interestingly, one of the plasma samples (subject No. 1 in Table 1) did not show any reactivity towards the FL-rFVIII, but only towards the BDD-rFVIII. The reason for this is not known and requires further study, but might be due to the flanking sequences introduced in the construct. The plasma of subject No. 15 showed an IgG-mediated, but unspecific binding to the FVIII molecule, as the binding was not possible to inhibit with excessive amounts of the molecule.