PB0133 - Identification of FVIII inhibitors by anti-FVIII ELISA assay and Bethesda assays in tandem to identify anti-Factor VIII inhibitors

Background: Anti-Factor VIII (FVIII) screening comprises of FVIII:C measurement and detection of anti-FVIII antibodies by ELISA and/or Bethesda assays. Nijmegen Bethesda is the standard assay for inhibitor detection with most factor replacements. Bovine-based reagents are used to assess for inhibitors in the presence of emicizumab. Porcine-based Bethesda assay measures FVIII inhibition with porcine FVIII therapy (e.g. obizur). Anti-FVIII ELISA detects both FVIII neutralising and non-neutralising antibodies and has been suggested to display greater sensitivity.

Aims: To compare matched samples for Bethesda and ELISA-based assays and assess sensitivity in detecting anti-FVIII inhibitors.

Methods: Samples were samples were assessed using routine laboratory methods for ELISA and Bethesda. The Bethesda assays used was dependent upon haemostatic factors replacement used.

Results: 218 anti-FVIII results were reviewed. 91% of patients showed correlation between ELISA and Bethesda assays whereas 9% did not. 5/218 (2%) patients had positive Nijmegen Bethesda with negative ELISA. 1/218 (0.5%) had a positive ELISA and negative Nijmegen Bethesda indicating a non-neutralising inhibitor with acquired haemophilia.

Table 1 and 2 show a false negative rate of 2% for anti-FVIII ELISA when compared to Nijmegen Bethesda. Rates of 8% false negatives with all Bethesda assay type if including 'borderline' positive Bethesda (FVIII recovery 62-90IU/dL and < 0.7BU/mL). These rates varied between the Bethesda assay used.

Conclusion(s): There was discrepancy between ELISA and Bethesda results in nearly 10% of patients tested. It is unclear if there is an anti-FVIII inhibitor present in these samples and suggests that employing the same cut-off values across Bethesda assay types may not be appropriate, particularly with 'borderline' results. We suggest that more detailed interpretation is required when using bovine- and porcine-based Bethesda assays.