Validation of 4% albumin as a diluent in the Bethesda Assay for FVIII inhibitors

Abstract

External quality assurance programs show the Nijmegen Bethesda Assay for FVIII inhibitors improves test specificity compared to the Classic Bethesda Assay but its uptake has been slow possibly due to the cost of using FVIII deficient plasma as diluent. This study was conducted to determine if modifying the Nijmegen Bethesda assay by replacement of FVIII deficient plasma with 4% as a diluent would be suitable for for measuring FVIII inhibitors. The titres of 59 samples from 35 patients with FVIII inhibitors were determined in parallel tests by the Nijmegen Bethesda Assay and and the modified Nijmegen assay. Method reproducibility was assessed on inhibitor-containing samples from seven individuals covering a range of titres from 1-200 Bethesda units/mL. The all-sample geometric mean titre was 6.73 Bethesda units/mL for the Nijmegen Bethesda Assay and 7.54 Bethesda units/mL for the modified Nijmegen assay. No sample was found where a difference in measured titre between methods would have altered clinical management. Agreement was very close in samples with titres less than 2BU/mL. Both assays gave inhibitor titres in external quality assurance samples of close to consensus values. The average between-run coefficients of variation were 8.6% for the Nijmegen Bethesda Assay and 7.9% for the modified Nijmegen assay. The modified Nijmegen assay using 4% albumin as the sample diluent showed good overall comparability to our existing Nijmegen Bethesda Assay and is substantially more cost-effective, making it a reasonable alternative for measuring FVIII inhibitors.