Discrepancies between ADAMTS13 activity assays in patients with thrombotic microangiopathies

ADAMTS13 activity assays are sometimes useful in confirming the clinical diagnosis or to distinguish different thrombotic microangiopathies (TMA). We investigated the commonly used clinical assays for ADAMTS13 activity. 159 samples from normal subjects or acquired TMA patients were studied in collagen binding (CBA), Fret and chromogenic peptide substrate assays. Frozen aliquots of pooled normal plasma gave similar values by CBA, Fret-VWF73 peptide, Fret-VWF86 and chromogenic VWF73 ELISA (chr-VWF73). Two lyophilised commercial calibrants gave lower ADAMTS13 activity by CBA than peptide substrate assays. The addition of solid HEPES to normal plasma caused a significant fall in CBA, but not Fret-VWF73 activity and might partly explain the differences, since lyophilised plasmas are often HEPES buffered. Normal plasmas showed good agreement between CBA and Fret assays, although chr-VWF73 gave slightly higher values. In acquired TMA, there was reasonable agreement between assays for samples with <11% ADAMTS13 activity (83% of samples showed agreement between CBA, Fret-VWF73 and chr-VWF73), but samples with moderate deficiency frequently showed lower CBA levels (only 41-52% agreement). However, there were also some discrepancies among the peptide substrate assays, with Fret-VWF86 sometimes giving slightly higher values than the VWF73 substrate assays. An International reference plasma might improve standardisation, but is not the only problem. It is unclear which assay has greatest clinical utility, this may depend on the nature of the sample. If the activity does not match the clinical picture, an alternative method should be performed. Where therapeutic monitoring is required, the same activity assay should be used throughout.