Diagnostic and prognostic value of factor VIII binding antibodies in acquired hemophilia A: data from the GTH-AH 01/2010 study

Background: Acquired hemophilia A is a severe bleeding disorder that requires fast and accurate diagnosis as it occurs often unexpectedly in previously healthy men and women of every age. The Nijmegen-modified Bethesda assay is the diagnostic reference standard for detecting neutralizing autoantibodies against factor VIII (FVIII), but is not widely available, not ideal for quantifying the complex type 2 inhibitors seen in acquired hemophilia, and suffers from high inter-laboratory variability. Objectives: To assess the diagnostic and prognostic value of FVIII-binding antibodies as detected by ELISA compared with the Nijmegen Bethesda assay. Methods: Samples from the time of first diagnosis and clinical data were available from 102 patients with acquired hemophilia enrolled in the prospective GTH-AH 01/2010 study. Controls (n = 102) were matched for gender and age. Diagnostic cut-offs were determined by receiver-operator curve analysis. The prognostic value was assessed in 92 of the 102 patients by Cox regression analysis of time to partial remission. Results: Anti-FVIII IgG above the 99th percentile (> 15 arbitrary units per mL) revealed high sensitivity and specificity (both 0.99; 95% confidence interval, 0.95-1.0) for diagnosing acquired hemophilia. The likelihood of achieving partial remission was related to anti-FVIII IgG concentration (< 300 arbitrary units, 1.0; 300-1050, 0.65; > 1050, 0.39). The Bethesda titer was only associated with the likelihood of partial remission when analyzed in the central laboratory, but not when data from local GTH study sites were used. Conclusion: Although the Nijmegen-modified Bethesda assay is the reference standard for demonstrating neutralizing antibodies, the detection of FVIII-binding antibodies by ELISA is similarly sensitive and specific for diagnosing acquired hemophilia. In addition, anti-FVIII IgG may provide prognostic information.