Measurement of B-domain-deleted ReFacto AF activity with a product-specific standard is affected by choice of reagent and patient-specific factors.

Abstract

Postinfusion ReFacto AF levels can be difficult to measure accurately due to discrepancies between one-stage and chromogenic FVIII assays. To overcome this, the use of the ReFacto AF laboratory standard (RAFLS) is recommended, but there are discordant reports regarding its usefulness. We investigated whether calibration with RAFLS and measurement of ReFacto AF levels are influenced by the choice of reagents and patient-specific factors in one-stage FVIII assays. Calibration curves were generated with both the RAFLS and a plasma standard using different F8DPs and one-stage FVIII assay reagents. This selection of reagents was then used to determine FVIII levels in the plasma of patients repeatedly treated with ReFacto AF. Results were compared with those obtained with a chromogenic assay. F8DP devoid of von Willebrand factor (VWF) falsely increased the values of RAFLS pro-coagulant activity generated using the APTT reagent. The resulting RAFLS calibration curve underestimated ReFacto AF levels to be half of their true concentration. The use of RAFLS with F8DP containing VWF reduced the discrepancy observed between the one-stage and chromogenic FVIII assays. However, the mean difference between the two assays still varied up to 50% depending on the patient. The RAFLS is a suitable calibrator for one-stage FVIII assays carried out with F8DP containing VWF. However, calibration with the RAFLS does not avoid the effect of patient-specific variables that contribute to discrepancies between the measurements of ReFacto AF levels with one-stage and chromogenic FVIII assays.