OBJECTIVETo evaluate the clinical value of modified Bethesda assay, modified Nijmegen assay and blank assay for detection of coagulation factor VIII (FVIII) inhibitors in patients with hemophilia A and analyze the factors that affect FVIII inhibitor detection.

METHODSThe levels of FVIII inhibitors in 257 patients with hemophilia A were detected using modified Bethesda assay, modified Nijmegen assay and blank assay (in which the buffer or FVIII-deficient plasma in the control mixture was replaced by deionized water). The 3 methods were compared for positivity rates and FVIII inhibitor titers based on the positive cut-off value of FVIII inhibitors ≥0.60 BU/mL.

RESULTSThe positive rates of modified Bethesda assay, modified Nijmegen assay and blank assay were 79.38%, 85.21% and 72.37%, respectively. A strong correlation was found between the results by modified Bethesda assay and modified Nijmegen assay (r=0.996, P<0.001), and FVIII inhibitor titers (P<0.001) but not the positive rates (P=0.105) detected by the two methods differed significantly. The correlation coefficients between modified Nijmegen assay and blank assay was 0.994 (P<0.001), and a significant difference was found in FVIII inhibitor titers (P<0.001) but not the positivity rates (P=0.079) detected by the two methods. The correlation coefficient between modified Nijmegen assay and blank assay was 0.994 (r=0.994), and the two methods yielded significantly different FVIII inhibitor titers and positivity rates (P=0.001).

CONCLUSIONThe modified Bethesda assay has a lower sensitivity than modified Nijmegen assay but has a higher sensitivity than blank assay. The consistency level of coagulation factors in the reaction system and stable buffer system are important factors that affect FVIII-inhibitor detection.