The challenge of identification of autoantibodies specific to systemic autoimmune rheumatic diseases in high throughput operation: Proposal of reliable and feasible strategies

• No one anti-ENA antibody assay provides satisfactory sensitivity and predictive value.
• Most standard commercial EIA assays have a low predictive value.
• Combining EIA and DID allows an efficient, reliable detection of anti-ENA antibodies.
• A positive predictive value can be improved by adjusting EIA assay cut-off values.

BackgroundAutoantibodies to extractable nuclear antigens (ENA) are good biomarkers for systemic autoimmune rheumatic diseases (SARD), but no one assay for the detection of these antibodies provides satisfactory sensitivity and positive predictive value (PPV). Here we evaluate current assays and propose novel strategies to detect anti-ENA antibodies.MethodsDiagnostic performance of double immunodiffusion (DID) and several enzyme immunoassays (EIA) for the detection of anti-ENA autoantibodies was determined using samples from 144 patients with a previous clinical diagnosis of SARD and 121 non-autoimmune individuals. A 2-step assay combining EIA and DID was developed and tested on 16,458 serum samples.ResultsEIA was more sensitive than DID for all anti-ENA antibodies, but yielded lower PPV (mean = 66%) than DID (mean = 96%) and a higher percentage of unexpected positive results. ROC-curve guided cut-off adjustments improved PPV for most EIA kits. Using the 2-step assay, over 80% of the samples were screened out by the first step (EIA), with results available within 24 h, leaving only about 20% to be confirmed by DID. 2.9% of the 16,485 samples were found to be positive.ConclusionsA 2-step assay combining the speed and potential for automation of EIA with the high specificity and PPV of DID allows efficient and reliable detection of anti-ENA antibodies. Alternatively, improved PPV can be achieved by adjusting cut-off values for EIA assay results.