Data shown in brackets are the percentages of the total number of samples  analysed.

IFA,  immunofluorescence assay.

The 103 samples were analysed using three commercially available ANA IFA kits, an ELISA and a multiplex assay called the BioPlex 2200. The following IFA kits were used: kit 1, ImmunoConcepts (distributed by GFMD, Novi, Michigan, USA); kit 2, Inova Diagnostics, San Diego, California, USA; and kit 3, Bio-Rad Kallestad (Bio-Rad, Hercules, California, USA). For the ImmunoConcepts kit, the Hep-2000 ANA-Ro kit was used. All assays were performed according to the manufacturer’s instructions, using secondary antibodies provided. For the IFA analysis, the serum samples were diluted 1:40 with 1× PBS to allow for the determination of whether the samples were positive or negative for ANA antibodies, as instructed by the manufacturer. Further titration of ANA-positive serum samples was not performed at this point of the studies. IFA was performed in one laboratory by two experienced observers, one of whom read kits 1 and 2; the other read kit 3. ANA-positive samples were defined by positive staining of the nucleus; staining of cytoplasm was not considered in this study in view of studies indicating the uncertain reliability of IFA in detecting antibodies to ribosomal P proteins (anti-P), a specificity that can lead to cytoplasmic staining.15 Only four sera had anti-P antibodies by the multiplex assays. Since these samples were all consistently ANA positive, our consideration of only nuclear staining appears to reasonably capture antibodies to relevant target antigens. The IFA slides were examined using the EVOS FL Cell Imaging System (Thermo Fisher, Waltham, Massachusetts, USA). An objective lens of ×20 was used, and the light source was an adjustable intensity LED. The samples also underwent ANA assessment by an ANA EIA as well as the BioPlex 2200 ANA Screen (both products of Bio-Rad); these assays were performed at Bio-Rad. The number (%) of samples identified as negative for each kit is shown. For IFA assays, indeterminate samples showed weak or borderline staining and could not be consistently classified as either negative or positive. Assays with modest elevations of anti-dsDNA are reported as indeterminate by the BioPlex 2200.