• Background/Purpose: The indirect immunofluorescence assay (IIF) is the gold standard for ANA testing.  However, many laboratories now use multiplex assay as an ANA screening test. Multiplex ANA assay tests for a limited number of autoantibodies (antibodies to dsDNA, chromatin, ribosomal P, SSA-52, SSA-60, SSB, Sm, SmRNP, RNP-A, RNP-68, Scl-70, Jo-1, centromere B) while IIF allows the detection of antibodies to a much larger number of autoantigens. There is a concern that reliance on multiplex analysis as a screening test might lead to cases where a positive ANA, and hence a diagnosis, is missed. Our study aims to evaluate the concordance between IIF and multiplex assay for ANA screening.
• Methods: ANA is tested by IIF at UVA. Samples are considered positive at a titer of 1:80. Thirteen specific autoantibodies are tested by an automated multiplex platform. For this study, samples with at least one autoantibody positive on multiplex analysis were considered ANA + by multiplex. A total of 545 blood samples sent to the UVA lab between June and November 2012 to be tested for ANA and specific autoantibodies were analyzed.
• Results: About 25.3% are ANA positive by IIF and 24.7% are ANA positive by multiplex. Only 13.8% of samples are positive by both methods (concordant). Kappa coefficient of agreement is 0.39 (fair). The most common antibody in all multiplex+ samples is that to RNP followed by SSA and dsDNA.  In the subset of multiplex + samples that were IIF negative, RNP was the most common antibody (40%). The most common pattern overall was homogenous. When the threshold for a positive ANA by IIF was increased to 1:160, the number of total samples that were IIF positive is reduced as expected (20.4%). Kappa coefficient of agreement was improved only to a value of 0.428 (moderate). A large number (74%) of the low titer (1:80) IFA assays were also multiplex negative. The majority (84.7%) of strongly positive IIF samples (1:640 and above) were also multiplex positive. However, 9 of the 59 high titer IIF samples were multiplex negative.
• Conclusion: We found a high rate of ANA positivity by both IIF and multiplex. Most of the discordant samples were either low titer IIF or had autoantibodies that might be less specific like anti-RNP. The agreement between the two tests as measured by a kappa co-efficient was only fair at a cutoff of 1:80 and improved to moderate at 1:160. This suggests that there is at best moderate agreement between the two tests. Also, in some cases, multiplex did not identify samples with high titer ANAs. We conclude that there is still a role for IIF to be used for ANA screening in patients suspected to have an autoimmune disease.

Fig 1: Concordance chart for ANA tested by IIF and Multiplex with a positive cutoff for IIF at a titer of 1:80.