LACK OF SPECIFICITY FOR ANTIBODIES TO DOUBLE-STRANDED DNA FOUND IN 4 COMMERCIAL KITS

Four commercially available kits (three enzyme linked immunosorbent assays and one modified Farr radioimmunoassay) were compared for their ability to detect specifically autoantibodies to double-stranded DNA (dsDNA) using 66 patient sera. This was assessed by comparing the results of the kits with those from an ELISA specifically measuring antibodies against highly purified dsDNA, single-stranded DNA (ssDNA), native DNA and histones. The RIA and two of the ELISAs seemed equally efficient at detecting antibodies to dsDNA, but all three also detected anti-ssDNA (the RIA being particularly bad for this). The need for highly purified dsDNA was clearly shown. The results obtained with one ELISA did not correlate with any variable investigated in this study. A total of 220 sera were assayed with the IDS RIA, of which 130 were recorded as positive. Of these, 50 sera seemed to contain no identifiable autoantibodies. This very high false positive rate may be due, at least in part, to precipitation of non-specifically bound labelled DNA.