Evaluation of a recombinant antigen enzyme-linked immunosorbent assay (ELISA) in the diagnostics of antinuclear antibodies (ANA) in children with rheumatic disorders

The reliability and accuracy of ELISAs for the detection of circulating ANA in children with rheumatic diseases has recently been questioned. In this study we evaluated an allegedly superior ELISA method using recombinant antigens in a paediatric population with various rheumatic conditions and compared it to a conventional Hep-2 IFA assay. Sera from 123 children (204 blood samples) were simultaneously tested by conventional ANA immunofluorescence on Hep-2 cells (ANA-IFA) and recombinant antigen ELISA (rELISA). There were 44 children with systemic lupus erythematosus (SLE), 29 with juvenile rheumatoid arthritis (JRA), eight with mixed connective tissue disease (MCTD), eight with reactive arthritis, five with juvenile fibromyalgia syndrome, three with dermatomyositis (JDMS) and 31 with other diagnoses.  Thirty-five children (27%) had a positive Hep-2 result, which remained undetected by ELISA (P <0.002). Almost all of these children had significant IFA titres above 1:160. The major discrepancies were observed in children with JRA and SLE. There was no titre correlation between the two assays and the rELISA’s OD readings were not linear.  The ELISA using recombinant antigens was not useful for the detection of serum ANA in children with rheumatic diseases due to a high rate of false negative results. These data concur with recent reports about the lack of reliability of ELISAs using non-recombinant antigens. Received: 2 May 2001 / Accepted: 10 September 2001