Aim: To compare the performance of a capture proteinase 3 enzyme linked immunosorbent assay (PR3 ELISA) with a direct PR3 ELISA in the measurement of PR3 antineutrophil cytoplasmic antibodies (ANCA).
Method: The performance of both assays systems was compared using two sets of sera. Sera from patients (n = 49) suffering from Wegener's granulomatosis (WG) and fulfilling the American College of Rheumatology classification criteria (or a modification of those criteria that allowed for ANCA positivity) were used along with sera from a group of patients (n = 48) considered to have a clinically false positive PR3 ANCA result when measured by routine direct ELISA.
Results: Using the assay specific cut-offs, the direct ELISA gave a positive result in 92% on repeat testing and the capture ELISA a positive result in 84% of sera from patients with WG. The capture ELISA was negative in 75% of patients considered to have a false positive PR3 ANCA on initial testing by direct ELISA (27% were negative on repeat testing by direct ELISA). The mean concentration of PR3 ANCA in WG patient sera measured by the capture ELISA was significantly higher than that measured by the direct ELISA. The capture PR3 ELISA had a broader analytical range which was also reflected in PR3 ANCA concentrations measured in serial serum samples from WG patients.
Conclusion: In this study the direct PR3 ELISA performed better as a screening test for PR3 ANCA compared with the capture PR3 ELISA, mainly because the cut-off for the capture ELISA needed to be set higher to avoid non-specific binding. In contrast, the improved analytical range of the capture ELISA made it a potentially more useful method for monitoring serial PR3 ANCA concentrations. In specific serum samples the capture ELISA was better able to discriminate 'false positive' PR3 ANCA.